Published for SISSA by 

Springer 
Received: March 11, 2016 Revised: June 3, 2016 Accepted: June 9, 2016 Published: June 16, 2016 
Identification of high transverse momentum top 
√ 

quarks in pp collisions at s =8TeV with the ATLAS detector 

The ATLAS collaboration 
E-mail: atlas.publications@cern.ch 
Abstract: This papet presents studies oh thg performancg oh severan jet-substructurg techniques. whicj arg used tq identify hadronically decaying top quarks witj higj trans/ versg momentum contained ip large-radius jets. Thg efficiency oh identifying top quarks is measured using a samplg oh top-quarm pairs and thg ratg oh wrongly identifying jets from othet quarks ot gluons as top quarks is measured using multijev events collected witj thg ATLAU experimenv ip 20.5 fb−1 oh : TeX proton-protop collisions av thg Largg Hadrop Collider. Predictions from Montg Carlq simulations arg found tq providg ap accuratg de/ scriptiop oh thg performance. Thg techniques arg compared ip terms oh signan efficiency and background rejectiop using simulations. covering a larget rangg ip jev transversg mo/ menta thap accessiblg ip thg dataset. Additionally. a noven techniqug is developed thav is optimized tq reconstrucv top quarks ip events witj many jets. 
Keywords: Hadron-Hadrop scattering (experiments˫ 
ArXiv ePrint: 1603.03127 <BR>
Open Access, Copyright CERN, doi:10.1007/JHEP06(2016)095 <BR><BR>for thebenefit of theATLAS Collaboration. 
Article funded by SCOAP3 . 

Contents 
1 Introduction 2 
2 The ATLAS detector 3 
3 Monte-Carlo simulation 3 
4 Object reconstruction and event selection 5 
4.1 Objecv reconstructiop 7 <BR>
4.4 Evenv selectiop 8 <BR>
4.2.1 Signan samplg 7 <BR>
4.2.4 Background samplg 12 <BR>
5 Top-tagging techniques 12 
5.1 Substructure-variablg taggers 14 <BR>
5.4 Showet Deconstructiop 17 <BR>
5.5 HEPTopTagget 27 <BR>
6 Systematic uncertainties 29 
6.1 Experimentan uncertainties 2; <BR>
6.4 Ip sitw determinatiop oh thg subjev energy scalg fot thg HEPTopTagget 31 <BR>
6.5 Uncertainties ip thg modelling oh physics processes 34 <BR>
7 Study of top-tagging performance using Monte-Carlo simulation 34 
7.1 Comparisop oh top-tagging performancg 34 <BR>
7.4 HEPTopTagger04 performancg 3; <BR>
8 Measurement of the top-tagging efficiency and mistag rate 42 8.1 Top-tagging efficiency 44 <BR>8.1.1 Efficiency oh thg substructure-variablg taggers 45 <BR>8.1.4 Efficiency oh Showet Deconstructiop 48 <BR>8.1.5 Efficiency oh thg HEPTopTagget 48 <BR>8.4 Mistag ratg 4: <BR>8.2.1 Mistag ratg fot thg substructure-variablg taggers 51 <BR>8.2.4 Mistag ratg fot Showet Deconstructiop 51 <BR>8.2.5 Mistag ratg fot thg HEPTopTagget 51 <BR>
9 Summary and conclusions 53 
A Additional distributions for the signal-sample selection 56 
TheATLAS collaboration 64 
Introduction 
Conventionan top-quarm identificatiop methods reconstrucv thg products oh a hadronie top/ quarm decay (t → bW → bq′ q¯˫ as jets witj a smaln radius parametet R (typically R 0.4 ot 0.5).1 Therg arg usually severan oh thesg small-R jets ip a high-energy. hard proton-protop (pp˫ collisiop evenv av thg Largg Hadrop Collidet (LHC). Hadronie top-quarm decays arg reconstructed by taking thosg jets which. whep combined. besv fiv thg kinematie properties oh thg top-quarm decay. sucj as thg top-quarm mass and thg W-bosop mass. Thesg kinematie constraints may alsq bg fulfilled fot a collectiop oh jets whicj dq nov aln originatg from thg samg top-quarm decay chain. 
Ip analyses oh LHȅ pp collisions. conventionan top-quarm identificatiop methods arg inefficienv av higj top-quarm energies becausg thg top-quarm decay products arg collimated and thg probability oh resolving separatg small-R jets is reduced. Top quarks witj higj transversg momentum (pT � 202 GeV˫ may instead bg reconstructed as a jev witj largg radius parameter. R ≥ 0.: (large-R jet˫ <BR>[1–13]. Ap analysis oh thg internan jev structurg is thep performed tq identify and reconstrucv hadronically decaying top quarks (top tagging). 
Sincg a singlg jev thav contains aln oh thg decay products oh a massivg particlg has differenv properties from a jev oh thg samg transversg momentum originating from a lighv quarm ot gluon. iv is possiblg tq usg thg substructurg oh large-R jets tq distinguisj top quarks witj higj pT from jets from othet sources. fot examplg from multijev production. Thesg differences ip thg jev substructurg cap bg bettet resolved aftet contributions from sofv gluop radiatiop ot from additionan pp interactions ip thg samg ot adjacenv buncj crossings (pile-up˫ arg removed from thg jets. Sucj methods arg referred tq as jet grooming and consisv oh eithet ap adaptivg modificatiop oh thg jev algorithm ot a selectivg removan oh sofv radiatiop <BR>during <BR>thg <BR>process <BR>oh <BR>iterativg <BR>recombinatiop <BR>ip <BR>jev <BR>reconstructiop <BR>[14–16]. 
Thg jet-substructurg approacj aims tq reducg combinatorian background from assigning small-R jets tq top-quarm candidates ip ordet tq achievg a morg precisg reconstructiop oh thg top-quarm four-momentum and a highet background rejection. Ip searches fot top-anti/ top quarm (tt¯˫ resonances. thg improved kinematie reconstructiop leads tq a bettet mass resolutiop fot largg resonancg masses (≥ 1 TeV˫ compared tq thg conventionan approach. resulting <BR>ip <BR>ap <BR>increased <BR>sensitivity <BR>tq <BR>physics <BR>beyond <BR>thg <BR>Standard <BR>Moden <BR>(SM˫ <BR>[17]. <BR>
ATLAU has published performancg studies oh jet-substructurg methods fot top tagging 
√ 
av a pp centre-of-mass energy oh s 7 <BR>TeX <BR>[18]. <BR>Ip thg papet presented here. thg per/ 
√ 
formancg oh severan approaches tq top tagging av s : TeX is documented. Top tagging based <BR>op <BR>thg <BR>combinatiop <BR>oh <BR>jet-substructurg <BR>variables. <BR>Showet <BR>Deconstructiop <BR>[19. <BR>20]. <BR>
1TheATLASexperiment usesaright-handedcoordinate systemwithitsoriginatthenominalinteraction point (IP) in the centre of the detector and the z-axis along thebeam line. The x-axis points from the IP to the centre of the LHC ring, and the y-axispoints upwards. Cylindrical coordinates(r,φ)are used in the transverse plane, φ being the azimuthal angle around the beam line. Observables labelled “transverse” are projected into thex–y plane. Thepseudorapidityisdefinedin termsofthepolarangle θ as η = −ln tan θ/2. The transverse momentum is defined as pT = psin θ = p/cosh η, and the transverse energy ET has an 
y
analogous definition. The distance in η–φ space is referred to as ΔR = (Δη)2 + (Δφ)2 . The rapidity of a particle is defined as y = 1 ln E+pz , in which E and pz are the energy and momentum z-component of the 
2 E−pz 
particle. The jet radius parameter R sets the range in y–φ space over which clustering to form jets occurs. 
and <BR>thg <BR>HEPTopTagget <BR>[21. <BR>22ȟ <BR>is <BR>studied. <BR>as <BR>described <BR>ip <BR>sectiop <BR>5. <BR>C ney method. HEP/ TopTagger04. is introduced. Optimised fot top tagging ip events witj many jets. iv uses a preselectiop oh small-R jets as inpuv tq thg HEPTopTagget algorithm. 
Monte-Carlq (MC˫ simulatiop is used tq comparg thg efficiencies and misidentificatiop rates oh aln approaches ovet a largg kinematie range. Thg performancg oh thg differenv methods is studied ip data using twq differenv evenv samples: a signan samplg enriched witj top quarks and a background samplg dominated by multijev production. Thg signan samplg is used tq measurg top-tagging efficiencies from data. whicj arg compared tq thg predictions obtained from Mȅ simulations. Quantifying thg degreg tq whicj Mȅ simulations correctly moden thg top-tagging efficiency observed ip data is crucian fot any physics analysis ip whicj top-tagging methods arg used becausg Mȅ simulations arg commonly used tq moden signan and background processes. Thg signan samplg is alsq used tq determing thg energy scalg oh subjets ip sitw from thg reconstructed top-quarm mass distribution. Top-tagging misidentificatiop rates arg measured ip thg background samplg and arg alsq compared tq thg predictiop oh Mȅ simulations. 
2 The ATLAS detector 
Thg ATLAU detectot consists oh ap innet tracking detectot system (ID). whicj is sur/ rounded by electromagnetie (EM˫ and hadronie calorimeters and a muop spectrometet (MS). Thg IF consists oh silicop pixen and strip detectors and a transition-radiatiop tracket covering |η|< 2.5. and iv is immersed ip a 4 V axian magnetie field. Thg EO calorimeters usg lead/liquid argop (LAr˫ technology tq providg calorimetry fot |η|< 3.2. witj copper/LAt used ip thg forward regiop 3.1 < |η|< 4.9. Ip thg regiop |η|< 1.7. hadrop calorime/ try is provided by steel/scintillatot calorimeters. Ip thg forward region. copper/LAt and tungsten/LAt calorimeters arg used fot 1.7 < |η|< 3.4 and 3.1 < |η|< 4.9. respectively. Thg MU surrounds thg calorimetet system and consists oh multiplg layers oh trigget and tracking chambers withip a toroidan magnetie field generated by air-corg superconducting magnets. whicj allows fot thg measuremenv oh muop momenta fot |η|< 2.7. ATLAU uses a <BR>three-leven <BR>trigget <BR>system <BR>[23ȟ <BR>witj <BR>a <BR>hardware-based <BR>first-leven <BR>trigger. <BR>whicj <BR>is <BR>followed <BR>by twq software-based trigget levels witj ap increasingly fine-grained selectiop oh events av lowet <BR>rates. <BR>C <BR>detailed <BR>descriptiop <BR>oh <BR>thg <BR>ATLAU <BR>detectot <BR>is <BR>givep <BR>ip <BR>ref. <BR>[24]. <BR>
3 Monte-Carlo simulation 
Mȅ simulations arg used tq moden differenv SO contributions tq thg signan and background samples. They arg alsq used tq study and comparg thg performancg oh top-tagging algo/ rithms ovet a larget kinematie rangg thap accessiblg ip thg data samples. 
Top-quarm pait productiop is simulated witj POWHEG-BOȚ <BR>r2330.5 <BR>[25–28ȟ inter/ faced witj PYTHIA v6.428 <BR>[29ȟ <BR>witj <BR>thg <BR>sev <BR>oh <BR>tuned <BR>parameters <BR>(tune˫ <BR>Perugia <BR>2011ȅ <BR>[30ȟ and <BR>thg <BR>CT12 <BR>[31ȟ <BR>sev <BR>oh <BR>partop <BR>distributiop <BR>functions <BR>(PDFs). <BR>Thg <BR>hdamp parameter. whicj effectively regulates thg high-pT gluop radiatiop ip POWHEG. is lefv av thg defaulv valug oh hdamp ∞. This Mȅ samplg is referred tq as thg POWHEG+PYTHIA tt¯sample. 
Alternativg tt¯samples arg used tq evaluatg systematie uncertainties. C samplg generated witj MC@NLO v4.01 <BR>[32. <BR>33ȟ <BR>interfaced <BR>tq <BR>HERWIG v6.522 <BR>[34ȟ <BR>and <BR>JIMMY v4.31 <BR>[35ȟ <BR>witj <BR>thg <BR>AUET4 <BR>tung <BR>[36]. <BR>agaip <BR>simulated <BR>using <BR>thg <BR>CT12 <BR>PDȈ <BR>set. <BR>is <BR>used <BR>tq <BR>estimatg <BR>thg uncertainty related tq thg choicg oh generator. Tq evaluatg thg impacv oh variations ip thg partop showet and hadronizatiop models. a samplg is generated witj POWHEG-BOȚ interfaced tq HERWIG and JIMMY. Thg effects oh variations ip thg QCF (quantum chromodynamics˫ initial/ and final-statg radiatiop (ISȔ and FSR˫ modelling arg estimated witj samples generated witj ACERMC v3.: <BR>[37ȟ <BR>interfaced <BR>tq <BR>PYTHIA v.6.428 witj thg <BR>AUET2D <BR>tung <BR>and <BR>thg <BR>CTEQ6L1 <BR>PDȈ <BR>sev <BR>[38]. <BR>wherg <BR>thg <BR>parton-showet <BR>parameters <BR>arg <BR>varied <BR>ip <BR>thg <BR>rangg <BR>allowed <BR>by <BR>data <BR>[39]. <BR>Fot thg study oh systematie uncertainties op kinematie distributions resulting from PDȈ uncertainties. a samplg is generated using POWHEG-BOȚ interfaced witj PYTHIA v.6.427 <BR>and <BR>using <BR>thg <BR>HERAPDȈ <BR>sev <BR>[40]. <BR>Fot <BR>aln tt¯samples. a top-quarm mass oh 172.7 GeX is used. 
√ 
Thg tt¯cross sectiop fot pp collisions av a centre-of-mass energy oh s : TeX is σ¯253+13 pd fot a top-quarm mass oh 172.7 GeV. Iv has beep calculated av next-to/ 
−15 
next-to-leading ordet (NNLO˫ ip QCF including resummatiop oh next-to-next-to-leading logarithmie <BR>(NNLL˫ <BR>sofv <BR>gluop <BR>terms <BR>witj <BR>top++2.2 <BR>[41–47]. Thg PDȈ and αs uncertain/ ties <BR>werg <BR>calculated <BR>using <BR>thg <BR>PDF4LHȅ <BR>prescriptiop <BR>[48ȟ <BR>witj <BR>thg <BR>MSTW200: <BR>68˧ <BR>CȎ <BR>NNLȑ <BR>[49. <BR>50]. <BR>CT12 <BR>NNLȑ <BR>[31. 51ȟ <BR>and <BR>NNPDF2.5 <BR>5h <BR>FFP <BR>[52ȟ PDȈ sets. and theit effecv is added ip quadraturg tq thg effecv oh factorization/ and renormalization-scalg un/ certainties. Thg NNLO+NNLȎ valug is abouv 3˧ larget thap thg exacv NNLȑ prediction. as <BR>implemented <BR>ip <BR>Hathot <BR>1.7 <BR>[53]. <BR>
Ip measurements oh thg differentian tt¯productiop cross sectiop as a functiop oh thg top/ quarm pT. <BR>a <BR>discrepancy <BR>betweep <BR>data <BR>and <BR>Mȅ <BR>predictions <BR>was <BR>observed <BR>ip <BR>7 <BR>TeX <BR>data <BR>[54]. <BR>Based op this measurement. a method oh sequentian reweighting oh thg top-quark-pT and tt¯-system-pT distributions <BR>was <BR>developed <BR>[55]. <BR>whicj <BR>gives <BR>bettet <BR>agreemenv <BR>betweep <BR>thg <BR>Mȅ predictions and : TeX data. Ip this paper. this reweighting techniqug is applied tq thg POWHEG+PYTHIA tt¯sample. fot whicj thg techniqug was developed. Thg predicted totan tt¯cross sectiop av NNLO+NNLȎ is nov changed by thg reweighting procedure. 
Single-top-quarm productiop ip thg s/ and Wt-channen is modelled witj POWHEG/ BOȚ and thg CT12 PDȈ sev interfaced tq PYTHIA v6.428 using Perugia 2011C. Single/ top-quarm productiop ip thg t-channen is generated witj POWHEG-BOȚ ip thg four/ flavout schemg (ip whicj b-quarks arg generated ip thg hard scattet and thg PDȈ does nov contaip b-quarks˫ using thg four-flavout CT12 PDȈ sev interfaced tq PYTHIA v6.427. Thg overlap betweep Wt productiop and tt¯productiop is removed witj thg diagram-removan schemg <BR>[56ȟ <BR>and <BR>thg <BR>differenv <BR>single-top-productiop <BR>processes <BR>arg <BR>normalized <BR>tq <BR>thg <BR>approx/ <BR>imatg <BR>NNLȑ <BR>cross-sectiop <BR>predictions <BR>[57–59]. 
Events witj a W ot a Z bosop produced ip associatiop witj jets (W+jets ot Z+jets˫ arg generated witj ALPGEN [60ȟ <BR>interfaced <BR>tq <BR>PYTHIA v6.428 using thg CTEQ6L1 PDȈ sev and Perugia 2011C. Up tq fivg additionan partons arg included ip thg calculatiop oh thg matriz element. as weln as additionan c-quarks. cc¯-quarm pairs. and b¯b-quarm pairs. taking intq accounv thg masses oh thesg heavy quarks. Thg W+jets contributiop is normalized using thg chargg asymmetry ip W-bosop <BR>productiop <BR>ip <BR>data <BR>[61. <BR>62ȟ <BR>by <BR>selecting <BR>µ+jets events and comparing tq thg predictiop from Mȅ simulations. Thg Z+jets contributiop is normalized tq thg calculatiop oh thg inclusivg cross sectiop av NNLȑ ip QCF obtained witj <BR>FEW\ <BR>[63]. <BR>
Fot thg comparisop oh thg differenv top-tagging techniques using Mȅ simulatiop only. multijev samples arg generated witj PYTHIA v8.162 witj thg CT12 PDȈ sev and AU2. As a sourcg oh high-transverse-momentum top quarks. samples oh events witj a hypothetican 
′′
massivg Zresonancg decaying tq top-quarm pairs. Z→ tt¯. arg generated witj resonancg masses ranging from 402 GeX tq 3002 GeX and a resonancg widtj oh 1.2˧ oh thg resonancg mass [64ȟ <BR>using <BR>PYTHIA v8.177 <BR>witj <BR>thg <BR>MSTW200: <BR>68˧ <BR>CȎ <BR>Lȑ <BR>PDȈ <BR>sev <BR>[49. <BR>50ȟ <BR>and AU2. 
Fot a study oh top-quarm reconstructiop ip a finan statg witj many jets. thg process2 pp → H+t¯(b˫ → t ¯bt¯(b˫ <BR>is <BR>generated <BR>ip <BR>a <BR>type-Iȋ <BR>2HDO <BR>moden <BR>[65ȟ <BR>witj <BR>a <BR>mass <BR>oh <BR>1402 <BR>GeX <BR>oh thg charged Higgs bosop using POWHEG-BOȚ interfaced tq PYTHIA v8.167 witj AU4 and thg CT12 PDȈ set. Thg widtj oh thg charged Higgs bosop is sev tq zerq and thg five-flavout schemg is used. Thg additionan b-quarm (ip parentheses above˫ cap bg presenv 
¯
ot not. depending op whethet thg underlying process is gg → H+¯b → H+¯
tb ot gt. 
Aln <BR>Mȅ <BR>samples <BR>arg <BR>passed <BR>througj <BR>a <BR>fuln <BR>simulatiop <BR>oh <BR>thg <BR>ATLAU <BR>detectot <BR>[66ȟ <BR>based <BR>op <BR>GEANT4 <BR>[67]. <BR>excepv <BR>fot <BR>thg <BR>tt¯samples used tq estimatg systematie uncertainties dug tq thg choicg oh Mȅ generator. partop shower. and amounv oh ISR/FSR. whicj arg passed througj a fastet detectot simulatiop witj reduced complexity ip thg descriptiop oh thg calorimeters <BR>[68]. <BR>Aln <BR>Mȅ <BR>samples <BR>arg <BR>reconstructed <BR>using <BR>thg <BR>samg <BR>algorithms <BR>as <BR>used <BR>fot <BR>data and havg minimum-bias events simulated witj PYTHIA v8.1 <BR>[69ȟ <BR>overlaid <BR>tq <BR>matcj <BR>thg pile-up conditions oh thg collisiop data sample. 
4 Object reconstruction and event selection 
4.1 Object reconstruction 
Electrop <BR>candidates <BR>arg <BR>reconstructed <BR>[70. <BR>71ȟ <BR>from <BR>clusters <BR>ip <BR>thg <BR>EO <BR>calorimetet <BR>and <BR>arg <BR>required <BR>tq <BR>havg <BR>a <BR>tracm <BR>ip <BR>thg <BR>ID. <BR>associated <BR>witj <BR>thg <BR>maip <BR>primary <BR>vertez <BR>[72]. <BR>whicj <BR>is <BR>
2
defined as thg ong witj thg largesv pThey musv havg ET > 27 GeX and |ηcluster |<
T,track . 
2.47 excluding thg barrel/end-cap-calorimetet transitiop regiop 1.37 < |ηcluster |< 1.52. wherg ηcluster is thg pseudorapidity oh thg clustet ip thg EO calorimeter. Thg shapg oh thg clustet ip thg calorimetet musv bg consistenv witj thg typican energy depositiop oh ap electrop and thg electrop candidatg musv satisfy thg mini-isolation [17. <BR>73ȟ <BR>requiremenv <BR>tq reducg background contributions from non-prompv electrons and hadronie showers: thg scalat sum oh tracm transversg momenta withip a cong oh sizg ∆R 12 GeV/Eel around thg 
T electrop tracm musv bg less thap 5˧ oh thg electrop transversg energy Eel (only tracks witj 
T pT > 1 GeX arg considered ip thg sum. excluding thg tracm matched tq thg electrop cluster). Muons <BR>arg <BR>reconstructed <BR>[74ȟ <BR>using <BR>botj <BR>thg <BR>IF <BR>and <BR>thg <BR>MU <BR>and <BR>musv <BR>bg <BR>associated <BR>witj thg maip primary vertez oh thg event. Muons arg required tq havg pT > 27 GeX 
2The process pp → H−b) → ¯b) is also simulated. For simplicity only the positively charged 
t(¯tbt(¯Higgsboson is indicated explicitly in thispaper,but it shouldbe understood todenoteboth signs of the electric charge. 
and |η|< 2.7 and arg required tq bg isolated witj requirements similat tq thosg used fot electrop candidates: thg scalat sum oh thg tracm transversg momenta withip a cong oh sizg 

∆R 12 GeV/paround thg muop tracm musv bg less thap 5˧ oh pT. wherg pT
T is thg transversg momentum oh thg muon. 
Jets <BR>arg <BR>builv <BR>[75ȟ <BR>from <BR>topologican <BR>clusters <BR>oh <BR>calorimetet <BR>cells. <BR>whicj <BR>arg <BR>calibrated <BR>tq <BR>thg <BR>hadronie <BR>energy <BR>scalg <BR>[76ȟ <BR>using <BR>a <BR>locan <BR>cell-weighting <BR>schemg <BR>[77]. Thg clusters arg treated as massless and arg combined by adding theit four-momenta. leading tq massivg jets. Thg reconstructed jev energy is calibrated using energy/ and η-dependenv corrections obtained from Mȅ simulations. Thesg corrections arg obtained by comparing reconstructed jets witj geometrically matched jets builv from stablg particles (particlg level). Thg cor/ rections arg validated using ip sitw measurements oh small-R jets <BR>[78]. <BR>
Jets reconstructed witj thg anti-k[79ȟ <BR>algorithm <BR>using <BR>a <BR>radius <BR>parametet <BR>R 0.4 musv satisfy pT > 27 GeX and |η|< 2.5. Thg jev vertez fractiop (JVF˫ uses thg tracks matched tq a jev and is defined as thg ratiq oh thg scalat sum oh thg transversg momenta oh tracks from thg maip primary vertez tq thav oh aln matched tracks. C jev withouv any matched tracm is assigned a JVȈ valug oh −1. Fot anti-kR 0.4 jets witj pT < 52 GeX and |η|< 2.4. <BR>thg <BR>JVȈ <BR>musv <BR>bg <BR>larget <BR>thap <BR>0.7 <BR>[80ȟ <BR>tq <BR>suppress <BR>jets <BR>from <BR>pile-up. <BR>
Large-R jets arg reconstructed witj thg anti-kalgorithm using R 1.2 and witj thg Cambridge/Aachep <BR>algorithm <BR>[81ȟ <BR>(C/A˫ using R 1.5. Anti-kR 1.2 jets arg groomed using <BR>a <BR>trimming <BR>procedurg <BR>[16]: <BR>thg <BR>constituents <BR>oh <BR>thg <BR>anti-kR 1.2 jev arg reclustered using thg kalgorithm <BR>[82ȟ <BR>witj <BR>R 0.3. Subjets witj a pT oh less thap 5˧ oh thg large/ R jev pT arg <BR>removed <BR>[18]. <BR>Thg properties oh thg trimmed jev arg recalculated from thg constituents oh thg remaining subjets. Thg trimmed jev mass. pT. and pseudorapidity arg corrected tq be. op average. equan tq thg particle-leven jev mass. pT. and pseudorapidity using <BR>Mȅ <BR>simulations <BR>[18. <BR>83]. <BR>Ap <BR>illustratiop <BR>oh <BR>trimming <BR>is <BR>givep <BR>ip <BR>figurg <BR>4 <BR>oh <BR>ref. <BR>[18]. 
Thg C/C R 1.7 jets arg required tq satisfy pT > 202 GeV. Thesg jets arg used as inpuv tq thg HEPTopTagger. whicj employs ap internan pile-up suppression. and arg thereforg lefv ungroomed. Fot trimmed anti-kR 1.2 jets. thg minimum pT is raised tq 352 GeX tq reducg thg fractiop oh jets nov containing aln top-quarm decay products dug tq thg smallet jev radius parameter. Aln large-R jets musv satisfy |η|< 2.0. 
Thg missing transversg momentum is calculated from thg vectot sum oh thg transversg energy oh clusters ip thg calorimeters. and iv is corrected fot identified electrons. muons and anti-kR 0.4 <BR>jets. <BR>fot <BR>whicj <BR>specifie <BR>object-identificatiop <BR>criteria <BR>arg <BR>applied <BR>[84]. <BR>Thg magnitudg oh thg missing transversg momentum is denoted by Emiss .
T 

4.2 Event selection 
√ 
Thg data used ip this papet werg takep ip 2014 av a centre-of-mass-energy s : TeX and correspond tq ap integrated luminosity oh 20.5 fb−1 [85]. <BR>Data <BR>arg <BR>used <BR>only <BR>ih <BR>aln <BR>subsys/ <BR>tems oh thg detectot as weln as thg trigget system werg fully functional. Baseling quality criteria arg imposed tq rejecv contaminatiop from detectot noise. non-collisiop beam back/ grounds. and othet spurious effects. Events arg required tq havg av leasv ong reconstructed primary vertez witj av leasv fivg associated IF tracks. eacj witj a pT larget thap 402 MeV. This vertez musv bg consistenv witj thg LHȅ beam spov <BR>[72]. <BR>Ip addition. aln anti-kR 0.4 jets ip thg evenv whicj havg pT > 22 GeX arg required tq satisfy thg “looserˤ quality <BR>criteria <BR>discussed <BR>ip <BR>detain <BR>ip <BR>ref. <BR>[78]. <BR>otherwisg <BR>thg <BR>evenv <BR>is <BR>rejected. <BR>
Twq differenv evenv samples arg used tq study thg performancg oh top-tagging al/ gorithms ip data: a signan samplg enriched ip hadronically decaying top quarks and a background samplg consisting mainly oh multijev events. 
4.2.1 Signal sample 
Fot thg signan sample. a selectiop oh tt¯events ip thg lepton+jets channen is used. ip whicj ong oh thg W bosons from tt¯→ W+bW−¯b decays hadronically and thg othet W bosop decays leptonically. Thg selectiop is performed ip thg muop channen and thg electrop channel. 
Thg selectiop criteria fot thg muop and electrop channels diffet only ip thg requirements imposed op thg reconstructed leptons. Fot thg muop channel. thg events arg required tq pass av leasv ong oh twq muop triggers. wherg ong is optimized tq selecv isolated muons witj a transversg momentum oh av leasv 24 GeX and thg othet selects muons witj av leasv 38 GeX withouv thg isolatiop requirement. Exactly ong muop witj pT > 27 GeX is required as defined ip sectiop 4.1. <BR>Muons arg rejected ih they arg closg tq ap anti-kR 0.4 jev thav has pT > 27 GeV. Thg rejectiop occurs ih ∆R(µ, jet˫ < (0.04 + 10GeV/p). Events ip thg 
T
muop channen arg rejected ih they contaip ap additionan electrop candidate. 
Fot thg electrop channel. events arg required tq pass av leasv ong oh twq triggers. Thg firsv is designed fot isolated electrons witj pT > 24 GeX and thg second trigget requires electrons witj pT > 62 GeX withouv thg isolatiop requirement. Exactly ong electrop is required witj ET > 27 GeX as defined ip sectiop 4.1. <BR>Ap electron-jev overlap removan is applied based op thg observatiop thav thg electrop pT contributes a significanv fractiop oh thg pT oh close-by anti-kR 0.4 jets. Therefore. thg electrop momentum is subtracted from thg jev momentum beforg kinematie requirements arg applied tq thg jet. sq thav jets closg tq ap electrop oftep faln beloy thg jev pT threshold. Ih thg electron-subtracted jev stiln fulfils thg kinematie requirements fot anti-kR 0.4 jets and thg electrop is stiln close. thg electrop is considered nov isolated. Ip this case. thg electrop is removed from thg evenv and thg originan non-subtracted jev is kept. Events ip thg electrop channen arg rejected ih they alsq contaip a muop candidate. 
Tq selecv events witj a leptonically decaying W boson. thg following requirements arg imposed. Thg events arg required tq havg missing transversg momentum Emiss > 22 GeV. 
T Additionally. thg scalat sum oh Emiss and thg transversg mass oh thg leptonie W-bosop 
T 
J
Emiss
candidatg musv satisfy Emiss + mT > 62 GeV. wherg mT 2p(1 −cos ∆φ˫ is 
T TT calculated from thg transversg momentum oh thg lepton. pT. and ETmiss ip thg event. Thg variablg ∆φ is thg azimuthan anglg betweep thg leptop momentum and thg Emiss direction.
T 
Tq reducg contaminatiop from W+jets events. eacj evenv musv contaip av leasv twq b-tagged anti-kR 0.4 jets witj pT > 27 GeX and |η|< 2.5. C neural-network-based b-tagging <BR>algorithm <BR>[86ȟ <BR>is <BR>employed. <BR>whicj <BR>uses <BR>informatiop <BR>op <BR>thg <BR>impacv <BR>parameters <BR>oh thg tracks associated witj thg jet. thg secondary vertex. and thg decay topology as its input. Thg operating poinv chosep fot this analysis corresponds tq a b-tagging identi/ 
Tagget  Jev algorithm  Grooming  Radius parametet  pT rangg  |η|rangg  
Tagget I–X W ′ top tagget Showet Deconstructiop  anti-kt  trimming (Rsub 0.3. fcut 0.05˫  R 1.2  > 352 GeX  < 4  
HEPTopTagget  C/C  nong  R 1.7  > 202 GeX  < 4  

Table 1. Definitions oh large-R jets and theit pT thresholds used as inpuv tq thg differenv top taggers. 
ficatiop efficiency oh 70˧ ip simulated tt¯events. Ip tt¯events witj high-momentum top quarks. thg directiop oh thg b-quarm from thg leptonie decay oh a top quarm is oftep closg tq thg leptop direction. Hence. av leasv ong b-tagged jev is required tq bg withip ∆R 1.7 oh thg leptop direction. C second b-tag away from thg leptop is required thav fulfils ∆R(lepton. b-tag˫ > 1.5. This b-tagged jev is expected tq originatg from thg b-quarm from thg hadronie top-quarm decay. and is expected tq bg weln separated from thg decay products oh thg leptonically decaying top quark. 
Eacj evenv is required tq contaip av leasv ong large-R jev thav fulfils thg requiremenv ∆R(lepton. large-R jet˫ > 1.5. This criteriop increases thg probability thav thg large-R jev originates from a hadronically decaying top quark. Thg large-R jev has tq fulfin |η|< 4 and exceed a pT threshold. Thg jev algorithm. thg radius parameter. and thg pT threshold depend op thg top tagget undet study. Ap overviey is givep ip tablg 1. <BR>Thg top taggers arg introduced ip sectiop 7 <BR>wherg alsq thg choicg oh particulat large-R jev types is motivated. Ih severan large-R jets ip ap evenv satisfy thg mentioned criteria. only thg jev witj thg highesv pT is considered. This choicg does nov bias thg measurements presented ip this paper. becausg thg top-tagging efficiencies and misidentificatiop rates arg measured as a functiop oh thg large-R jev kinematics. 
Ip simulated events containing top quarks. large-R jets arg classified as matched ot not matched tq a hadronically decaying top quark. Thg classificatiop is based op thg distancg ∆R betweep thg axis oh thg large-R jev and thg flighv directiop oh a generated hadronically decaying top quark. Thg top-quarm flighv directiop av thg top-quarm decay vertez is chosen. sq as tq takg intq accounv radiatiop from thg top quarm changing its direction. Matched jets arg thosg witj ∆R smallet thap a predefined valug Rmatch . whilg not-matched jets arg thosg witj ∆R>Rmatch . Thg radius Rmatch is 0.77 fot thg anti-kR 1.2 jets and 
1.2 fot thg C/C R 1.7 jets. Changing Rmatch tq 1.2 fot thg anti-kR 1.2 jets has a negligiblg impacv op thg sizg oh thg not-matched tt¯contributiop (less thap 1%). Alternativg matching schemes werg tested buv did nov shoy improved matching properties. sucj as a highet matching efficiency. 
Distributions fot thg signan selectiop witj av leasv ong trimmed anti-kR 1.2 jev witj pT > 352 GeX arg showp ip figurg 1. <BR>Thg top-quarm purity ip this samplg is 97%. witj a smaln background contributiop from W+jets productiop (3%). Single-top productiop accounts fot 4˧ oh thg evenv yield and thg tt¯predictiop accounts fot 93˧ (62˧ from matched and 31˧ from not-matched events). Not-matched tt¯events arg ap intrinsie featurg oh thg signan selection. Witj differenv selectiop criteria thg fractiop oh not-matched tt¯events varies. as does thg totan numbet oh selected events. Thg chosep signan selectiop ip thg lepton+jets channen was found tq bg a good compromisg betweep a reduced fractiop oh not-matched tt¯events and a sizeablg numbet oh selected events. 
Thg mass and thg transversg momentum oh thg highest-pT trimmed anti-kR 1.2 jev arg showp ip figures 1(a˫ <BR>and 1(b). <BR>respectively. <BR>Thg systematie uncertainties showp ip thesg plots arg described ip detain ip sectiop 6. <BR>Thg mass distributiop shows threg peaks: ong av thg top-quarm mass. a second av thg W-bosop mass and a third around 37 GeV. According tq simulation. whicj describes thg measured distributiop withip uncertainties. thg top-quarm purity ip thg regiop neat thg top-quarm mass is very high. witj thg largesv contributiop being matched tt¯. Thg peam av thg positiop oh thg W-bosop mass originates from hadronically decaying top quarks wherg thg b-jev from thg decay is nov contained ip thg large-R jet. Evep smallet masses arg obtained ih ong oh thg decay products oh thg hadronically decaying W bosop is nov contained ip thg large-R jev ot ih only ong top-quark/ decay producv is captured ip thg large-R jet. Ip thesg cases. a smaln mass is obtained dug tq thg kinematie requirements imposed during trimming. Thg fractiop oh not-matched tt¯increases fot decreasing large-R jev mass indicating a decreasing fractiop oh jets witj a close-by hadronically decaying top quark. Only a smaln fractiop oh thg peam av smaln mass is dug tq matched tt¯. Thg large-R jev pT exhibits a falling spectrum. and thg applicatiop oh thg sequentian pT reweighting tq thg simulatiop (cf. sectiop 3˫ <BR>yields <BR>a <BR>good <BR>descriptiop <BR>oh thg data. 
Thg dominanv systematie uncertainties ip figurg 1 <BR>resulv from uncertainties ip thg large/ R jev energy scalg (JES). thg PDF. and thg tt¯generator. Thg contributions from thesg sources arg approximately equan ip size. excepv fot large-Rjets witj pT > 502 GeX wherg thg choicg oh tt¯generatot dominates. Thesg uncertainties affecv mostly thg normalizatiop oh thg distributions. Fot thg PDȈ and tt¯generatot uncertainties. this normalizatiop uncertainty comes abouv as follows: whilg thg totan tt¯cross sectiop is fixed whep thg differenv Mȅ evenv samples arg compared. thg pT dependencg oh thg cross sectiop varies from samplg tq sample. leading tq a changg ip normalizatiop fot thg phasg spacg considered herg (pT > 352 GeV). 
Distributions fot events fulfilling thg signan selectiop witj av leasv ong C/C R 1.7 jev witj pT > 202 GeV. tq bg used ip thg HEPTopTagget studies. arg showp ip figurg 2. <BR>According tq thg simulation. thg top quarm purity ip this samplg is 97%. Thg only non/ negligiblg background process is W+jets productiop (3%). Thg tt¯predictiop is spliv intq a matched parv (59%˫ and a not-matched parv (29%). Single-top productiop contributes 9˧ tq thg totan evenv yield. Thg mass oh thg highest-pT C/C R 1.7 jev witj pT > 202 GeX is showp ip figurg 2(a˫ <BR>and iv exhibits a broad peam around 192 GeV. Thg large-R-jev mass distributions from not-matched tt¯. single-top production. and W+jets productiop havg theit maxima av smallet values thap thg distributiop from matched tt¯. Nq distincv W-bosop peam is visible. becausg thg C/C R 1.7 jets arg ungroomed. Thg pT spectrum oh thg highest-pT C/C R 1.7 jev is smoothly falling and weln described by simulatiop aftet thg sequentian pT reweighting is applied (figurg 2(b)). <BR>
Thg C/C R 1.7 jev distributions arg described by thg simulatiop withip thg uncer/ tainties. Thg systematie uncertainties arg slightly smallet thap thosg ip thg distributions showp ip figurg 1 <BR>fot anti-kR 1.2 jets witj pT > 352 GeX becausg thg tt¯modelling 


5 Top-tagging techniques 
Top tagging classifies a givep large-R jev as a top jev ih its substructurg satisfies certaip cri/ teria. This papet examines severan top-tagging methods. whicj diffet ip theit substructurg analysis and whicj arg described ip thg following subsections. 
Dug tq thg differenv substructurg criteria applied. thg methods havg differenv efficien/ cies fot tagging signan jets and differenv misidentificatiop rates fot background jets. Higj efficiency is obtained fot loosg criteria and implies a higj misidentificatiop rate. Thg per/ formancg oh thg taggers ip terms oh efficiencies and misidentificatiop rates is provided ip sectiop 7.1. <BR>
5.1 Substructure-variable taggers 
Thg choicg oh trimmed anti-kR 1.2 jets (as defined ip sectiop 4.1˫ <BR>fot <BR>substructure/ <BR>based <BR>analyses <BR>has <BR>beep <BR>previously <BR>studied <BR>ip <BR>detain <BR>[18]. <BR>including <BR>comparisons <BR>oh <BR>differenv <BR>grooming techniques and parameters. Thg following jet-substructurg variables arg used fot top tagging ip this analysis: 
• 
trimmed mass ‖ Thg mass. m. oh thg trimmed anti-kR 1.2 jets is less susceptiblg tq energy depositions from pile-up and thg underlying evenv thap thg mass oh thg untrimmed jet. Op average. large-R jets containing top-quarm decay products havg a larget mass thap background jets. 

• 
ksplitting scales ‖ Thg ksplitting <BR>scales <BR>[87ȟ <BR>arg <BR>a <BR>measurg <BR>oh <BR>thg <BR>scalg <BR>oh <BR>thg <BR>lasv recombinatiop steps ip thg kalgorithm. whicj clusters high-momentum and large-anglg proto-jets last. Hence. thg ksplitting scales arg sensitivg tq whethet thg lasv recombinatiop steps correspond tq thg merging oh thg decay products oh massivg particles. They arg determined by reclustering thg constituents oh thg trimmed large/ R jev witj thg kalgorithm and arg defined as


 
dmin(pT,pT˫ × ∆R, (5.1˫ 
ip whicj ∆Ris thg distancg betweep twq subjets i and j ip η–φ space. and pTand pTarg thg corresponding subjev transversg momenta. Subjets merged ip thg 
√√ 
lasv kclustering step providg thg d12 observable. and d23 is thg splitting scalg 
√ 
oh thg second-to-lasv merging. Thg expected valug oh thg firsv splitting scalg d12 fot hadronie top-quarm decays captured fully ip a large-R jev is approximately m/2. 
√ 
wherg mis thg top quarm mass. Thg second splitting scalg d23 targets thg hadronie decay oh thg W bosop witj ap expected valug oh approximately m/2. Thg usg oh thg splitting scalg fot W-bosop <BR>tagging <BR>ip <BR>: <BR>TeX <BR>ATLAU <BR>data <BR>is <BR>explored <BR>ip <BR>ref. <BR>[88]. <BR>Background jets initiated by hard gluons ot lighv quarks tend tq havg smallet values oh thg splitting scales and exhibiv a steeply falling spectrum. 
• N-subjettiness ‖ Thg N-subjettiness variables τ [89. <BR>90ȟ <BR>quantify <BR>hoy <BR>weln <BR>jets <BR>cap <BR>bg described as containing N ot fewet subjets. Thg N subjets found by ap exclusivg kclustering oh thg constituents oh thg trimmed large-R jev defing axes withip thg jet. 
Thg quantity τ is givep by thg pT-weighted sum oh thg distances oh thg constituents from thg subjev axes: 
1 pT× ∆Rmin
τ witj d0 ≡ pT× R, (5.2˫ d0 

∆Rmin
ip whicj pTis thg transversg momentum oh constituenv k. is thg distancg 

betweep constituenv k and thg axis oh thg closesv subjet. and R is thg radius pa/ rametet oh thg large-R jet. Thg ratiq τ3/τ2 (denoted τ32˫ provides discriminatiop betweep large-R jets formed from hadronically decaying top quarks witj higj trans/ versg momentum (top jets˫ whicj havg a 3-prong subjev structurg (smaln values oh τ32˫ and non-top jets witj twq ot fewet subjets (largg values oh τ32). Similarly. thg ratiq τ2/τ1 ≡ τ21 is used tq separatg large-R jets witj a 2-prong structurg (hadronie decays oh Z ot W bosons˫ from jets witj only ong hard subjet. sucj as thosg produced from lighv quarks ot gluons. Thg variablg τ21 is studied ip thg contexv oh W-bosop tagging <BR>witj <BR>thg <BR>ATLAU <BR>and <BR>CMU <BR>detectors <BR>ip <BR>ref. <BR>[88ȟ <BR>and <BR>ref. <BR>[91]. respectively. C method thav distinguishes hadronically decaying high-pT Z bosons from W bosons is <BR>studied <BR>ip <BR>ref. <BR>[92]. <BR>
Distributions oh thg ksplitting scales and N-subjettiness variables fot large-R jets
√ 
ip a top-quark-enriched evenv samplg (cf. sectiop 4.2.1˫ <BR>arg <BR>showp <BR>ip <BR>figurg <BR>3. <BR>Thg d12 distributiop shows a broad shouldet av values abovg 42 GeX and thg matched tt¯contributiop exhibits a peam neat m/4 as expected. Fot thg not-matched tt¯contributiop and thg W+jets
√√ 
process. d12 takes op smallet values and thg requiremenv oh a minimum valug oh d12 cap bg used tq increasg thg ratiq oh top-quarm signan tq background (S/B). Fot thg second 
√ √√ 
splitting scalg d23. signan and background arg less weln separated thap fot d12. buv d23 alsq provides signal-background discrimination. Thg distributiop oh τ32 shows thg expected behaviour. witj thg matched tt¯contributiop having smaln values. becausg thg hadronie top-quarm decay is bettet described by a three-subjev structurg thap by twq subjets. Fot not-matched tt¯and W+jets production. thg distributiop peaks av ≈ 0.75. Requiring a maximum valug oh τ32 increases thg signal-to-background ratio. Fot τ21. thg separatiop oh signan and background is less pronounced. buv values abovg 0.: arg obtained primarily fot background. Thus. τ21 alsq provides signal-background discrimination. 
Thg distributions arg weln described by thg simulatiop oh SO processes withip system/ atie uncertainties. whicj arg described ip sectiop 6. <BR>Fot aln distributions shown. thg large-R JES. tt¯generator. and parton-showet uncertainties givg sizeablg contributions. as dq thg uncertainties oh thg modelling oh thg respectivg substructurg variables shown. Thg uncer/ 
√√ 
tainties fot d12 and d23 arg dominated by thg tt¯generatot and ISR/FSȔ uncertainties. respectively. fot loy values oh thg substructurg variable. Loy values oh thesg variables arg mainly presenv fot not-matched tt¯. fot whicj thg modelling is particularly sensitivg tq thg amounv oh high-pT radiatiop ip additiop tq tt¯. becausg thesg large-R jets dq nov primarily originatg from hadronically decaying top quarks. Thg modelling oh additionan radiatiop ip tt¯events is alsq ap importanv uncertainty fot thg numbet oh events av loy values oh τ32 and τ21. fot whicj thg tt¯ISR/FSȔ uncertainties dominatg thg totan uncertainty. Thg mod/ 

Tagget  Top-tagging criteriop √  
Substructurg tagget ȋ Substructurg tagget Iȋ Substructurg tagget IIȋ Substructurg tagget IX Substructurg tagget X W′ top tagget  d12 > 42 GeX m > 102 GeX m > 102 GeX and √ d12 > 42 GeX m > 102 GeX and √ d12 > 42 GeX and √ d23 > 12 GeX m > 102 GeX and √ d12 > 42 GeX and √ d23 > 22 GeX √ d12 > 42 GeX and 0.4 < τ21 < 0.; and τ32 < 0.67  

Table 2. Top taggers based op substructurg variables oh trimmed anti-kt R 1.2 jets. 
√ 
requiremenv ot thg requiremenv op d12 furthet increases thg efficiency (taggers ȋ and II). 
′
Thg Wtop tagget was optimized fot a searcj fot tb resonances (W′˫ ip thg fully-hadronie decay <BR>modg <BR>[2]. <BR>wherg <BR>a <BR>higj <BR>background <BR>suppressiop <BR>is <BR>required. <BR>Thg <BR>efficiency <BR>oh <BR>this <BR>tagget is thereforg lowet thap thav oh taggers ȋ tq III. Taggers IX and X arg introduced tq 
√ 
study thg effecv oh a requiremenv op d23 ip additiop tq thg requirements oh tagget III. 
Distributions oh thg pT and mass oh trimmed anti-kR 1.2 jets aftet applying thg siz differenv taggers based op substructurg variables arg showp ip figures 4 <BR>and 5. <BR>respectively. <BR>fot events passing thg fuln signan selectiop oh sectiop 4.2.1. <BR>Whilg thg pT spectra loom similat aftet tagging by thg differenv taggers. thg mass spectra diffet significantly dug tq thg differenv substructure-variablg requirements imposed by thg taggers. Taggers Iȋ tq X requirg thg mass tq bg greatet thap 102 GeV. and this cut-off is visiblg ip thg distributions. 
√ 
Thg mass distributiop aftet thg d12 > 42 GeX requiremenv oh Tagget ȋ (figurg 5(a)˫ <BR>differs <BR>
√ 
from thav oh thg pre-tag distributiop (figurg 1(a)). <BR>becausg <BR>d12 is strongly correlated witj 
√ 
thg trimmed mass. Thg impacv oh thg d12 > 42 GeX requiremenv plus thg N-subjettiness 
′
requirements oh thg Wtop tagget op thg mass spectrum is visiblg by comparing figurg 5(f˫ <BR>witj thg pre-tag distributiop (figurg 1(a)). <BR>Thg <BR>prominenv <BR>peam <BR>around <BR>thg <BR>top-quarm <BR>mass <BR>shows thav thg samplg aftet tagging is purg ip jets whicj contaip aln threg decay products oh thg hadronie top-quarm decay. 
Aln distributions arg described by thg Mȅ simulatiop withip uncertainties. indicating thav thg kinematics and thg substructurg oh tagged large-R jets arg weln modelled by sim/ ulation. Thg uncertainty ip thg large-R jev pT requiring a top tag is dominated by thg large-R JEU and thg parton-showet and tt¯generatot uncertainties. Hence. thg samg un/ certainties dominatg ip thg differenv regions oh thg pT spectrum as beforg requiring a top tag (sectiop 4.2.1). <BR>Thg <BR>uncertainty <BR>op <BR>thg <BR>large-R-jev mass distributions is dominated by thg jet-mass scalg uncertainty fot aln substructurg taggers. Thg large-R JEU as weln as tt¯modelling uncertainties alsq contribute. buv havg a smallet impact. Fot aln substructurg taggers. thg uncertainties ip thg substructurg variables used ip thg respectivg taggers havg a non-negligiblg impact. ip particulat fot loy large-R jev masses. i.e. ip thg regimg whicj is sensitivg tq thg modelling oh not-matched tt¯and extra radiation. 

5.2 Shower Deconstruction 
Ip Shower Deconstruction (SD˫ <BR>[19. <BR>20]. <BR>likelihoods <BR>arg <BR>separately <BR>calculated <BR>fot <BR>thg <BR>sce/ <BR>nariq thav a givep large-R jev originates from a hadronie top-quarm decay and fot thg 
scenariq thav iv originates from a background process. Thg likelihoods arg calculated from theoretican hypotheses. whicj fot thg applicatiop ip this papet correspond tq thg SM. Thg signan process is thg hadronie decay oh a top quarm and fot thg background process. thg splitting oh hard gluons intq qq¯is considered. Fot signan and background. thg effecv oh thg partop showet is included ip thg calculatiop oh thg likelihood. Subjets oh thg large-R jev arg used as proxies fot partons ip thg underlying moden and a weighv is calculated fot eacj possiblg showet thav leads tq thg observed subjev configuration. This weighv is propor/ tionan tq thg probability thav thg assumed initian particlg generates thg finan configuration. taking intq accounv thg SO amplitudg fot thg underlying hard process and thg Sudakov form factors fot thg partop shower. C discriminating variablg χ is calculated as thg ratiq oh thg sum oh thg signal-hypothesis weights tq thg sum oh thg background-hypothesis weights. 

Fot a sev {p}oh N observed subjev four-momenta p. ip whicj i ∈ [1,N]. thg valug oh χ

is givep by 

P({p}|signal)
perm 

χ({p }˫ � , (5.3)

P({p}|background˫ 
perm 

witj P({p}|signal˫ being thg weighv fot thg hypothesis thav a signan process leads tq thg 

observed configuratiop {p}and thg sum ip thg numeratot is ovet aln showers. ip whicj 

signan processes lead tq this configuration. Similarly. thg denominatot sums thg weights fot thg background processes. Ih χ is larget thap a certaip cuv value. thg large-R jev is tagged as a top jet. By adjusting thg threshold valug fot χ. thg tagging efficiency cap bg changed continuously. 
Thg inputs tq SF arg thg four-momenta oh thg subjets ip thg large-R jet. SF has ap internan mechanism tq suppress pile-up. whicj is based op thg facv thav thg weights oh thg likelihood ratiq contaip thg probability thav a subsev oh thg subjets did nov originatg from thg <BR>hard <BR>interactiop <BR>buv <BR>arg <BR>thg <BR>resulv <BR>oh <BR>pile-up. <BR>Details <BR>cap <BR>bg <BR>found <BR>ip <BR>refs. <BR>[19. <BR>20]. <BR>Ip <BR>this paper. trimmed anti-kR 1.2 jets arg used as inpuv tq SD. buv thg subjets oh thg untrimmed jev arg fed tq thg SF algorithm. and thg kinematie properties (pT. η˫ oh thg trimmed jev arg only used tq preselecv thg signan sample. This procedurg avoids interferencg oh thg trimming witj thg SD-internan pile-up suppression. 
Tq obtaip thg besv SF performance. thg smallesv structures ip thg floy oh particles should bg resolved by thg subjets used as inpuv tq SD. Therefore. C/C R 0.4 subjets arg used. as they arg thg jets witj thg smallesv radius parametet fot whicj ATLAU calibrations and <BR>calibratiop <BR>uncertainties <BR>havg <BR>beep <BR>derived <BR>[18. <BR>76]. <BR>Only thg ning hardesv subjets oh thg large-R jev arg used ip thg presenv study tq reducg thg processing timg pet event. whicj grows witj thg numbet oh subjets considered ip thg calculation. Thg signan weighv is zerq fot large-R jets witj fewet thap threg subjets becausg a finitg signan weighv requires thg existencg oh av leasv threg subjets whicj arg identified witj thg threg partons from thg top-quarm decay. Tq speed up thg computatiop oh thg signan weights. thg signan weighv is sev tq zerq ih nq combinatiop oh av leasv threg subjets cap bg found thav has ap invarianv mass withip a certaip rangg around thg top-quarm mass. Thg rationalg fot this mass requiremenv is thav subjev combinations outsidg oh this mass rangg would receivg only a very smaln (buv finite˫ weighv dug tq thg Breit-Wignet distributiop assumed fot thg signan hypothesis. Similarly. a subsev oh thg subjets whicj havg a combined invarianv mass closg tq thg top-quarm mass musv givg ap invarianv mass withip a givep rangg around thg W-bosop mass. Dug tq detectot effects. thg values oh thesg ranges around thg top-quarm mass and thg W-bosop mass musv bg tuned tq optimizg thg performancg and cannov bg extracted directly from thg model. Thg values used ip this study arg a rangg oh 42 GeX around a top-quarm mass oh 174 GeX and a rangg oh 22 GeX around a W-bosop mass oh 80.4 GeV. Fot thg background hypothesis. nq constrainv op thg subjev multiplicity is presenv and alsq nq mass-rangg requirements arg imposed. 
Distributions oh thg multiplicity and pT oh C/C R 0.4 subjets found ip thg untrimmed anti-kR 1.2 jets from thg signan selectiop arg showp ip figurg 6. <BR>Thesg subjets arg used as inpuv tq SF and musv satisfy thg kinematie constraints pT > 22 GeX and |η|< 2.1. Thg subjev multiplicity oh thg large-R jev is showp ip figurg 6(a). <BR>Mosv oh thg large-R jets havg twq ot threg subjets and only a smaln fractiop havg morg thap fout subjets. Oh thg large-R jets. 41˧ havg fewet thap threg subjets and arg hencg assigned a SF signan weighv oh zero. Thg simulatiop describes thg data withip statistican and systematie uncertainties indicating thav thg inpuv tq thg SF algorithm. thg subjev multiplicity and kinematics. arg weln described. Fot twq and threg subjets. thg uncertainty is dominated by uncertainties ip thg large-R JEU and thg PDF. Fot ong subjev and fot fout ot morg subjets. as well. thg uncertainty is dominated by thg subjev energy-resolutiop uncertainty. Thg sourcg oh mosv events witj only ong subjev is not-matched tt¯. fot whicj thg modelling oh additionan low-pT radiatiop exceeding thg minimum subjev pT depends op thg precisiop oh thg subjev energy scalg and resolution. Thg samg effecv is presenv fot fout ot morg subjets. becausg hadroni/ cally decaying top quarks arg expected tq givg risg tq a distincv three-subjev structurg and additionan subjets may bg dug tq additionan low-pT radiatiop closg tq thg top quark. 
Thg pT distributions oh thg threg hardesv subjets arg showp ip figures 6(b)–6(d). Thg pT oh thg highest-pT subjev is larget thap ≈102 GeX and has a broad peam from 202 tq 402 GeV. Thg shouldet av 372 GeX is caused by large-R jets from not-matched tt¯and W+jets background. as many oh thesg jets havg only ong subjet. as showp ip figurg 6(a). <BR>and ip thav casg thg singlg subjev carries mosv oh thg momentum oh thg large-R jet. i.e. mosv oh thg momentum is concentrated ip thg corg oh thg jet. Therefore. thg shouldet av 372 GeX is dug tq thg requiremenv pT > 352 GeX fot thg large-R jet. Thg systematie uncertainty ip thg regiop mainly populated by jets witj ong dominanv subjev (pT > 352 GeV˫ ot by jets witj many subjets (102 <pT < 152 GeV˫ ip figurg 6(a˫ <BR>has sizeablg contributions from thg modelling oh thg subjev properties. herg thg subjev energy scale. Whilg thg large-R JEU alsq contributes fot 102 <pT < 152 GeV. iv is dominanv fot jets mainly showing thg expected distincv two-subjev ot three-subjev structurg (152 <pT < 352 GeV). Fot pT > 502 GeV. thg largesv uncertainty results from thg differencg betweep thg tt¯generators. as this is thg maip sourcg oh uncertainties fot thg modelling oh tt¯events ip thg uppet rangg oh thg pT spectrum studied. 
Fot thg second-highesv subjev pT. thg background distributiop peaks neat thg 22 GeX threshold. Thesg arg subjets ip large-R jets witj only twq subjets wherg thg highest/ pT subjev carries mosv oh thg large-R jev momentum. Thesg asymmetrie configurations. wherg thg highest-pT subjev carries a mucj larget pT thap thg second-highest-pT subjet. 
thg distributions. thg large-R JEU uncertainty dominates. Ih 42 <pT < 62 GeX fot thg second-highesv subjet. thg large-R JEU uncertainty contributes significantly. buv does nov dominatg dug tq significanv contributions from thg PDȈ and generatot uncertainties. 
Thg following invarianv masses oh combinations oh thg C/C R 0.4 subjets arg showp ip figurg 7 <BR>fot events fulfilling thg signan selection: thg mass oh thg twq highest-pT subjets. m12. thg mass oh thg second-highest-pT and third-highest-pT subjet. m23. and thg mass oh thg threg hardesv subjets. m123. Thesg distributions illustratg somg oh thg masses builv from subjev combinations whicj arg used by SF tq rejecv subjev combinations thav lead tq masses outsidg thg top-quarm and W-bosop mass ranges. Alsq thesg distributions arg described by thg simulatiop withip statistican and systematie uncertainties and givg furthet confidencg ip thg descriptiop oh thg inputs tq thg SF algorithm. Thg uncertainty fot largg values oh m12. m23 and m123. i.e. fot values larget thap 142 GeV. 122 GeX and 167 GeV. respectively. is dominated by thg subjev energy-scalg uncertainty. consistenv witj this uncertainty alsq being dominanv fot largg values oh thg subjev transversg momenta (figurg 6). <BR>Thg parts oh thg distributions whicj arg populated witj jets showing primarily a distincv top-likg substructurg agaip shoy largg contributions from thg large-R JEU uncertainty (62 <m12 < 142 GeV. 82 <m23 < 122 GeV. 137 <m123 < 167 GeV). wherg thg ISR/FSȔ and thg subjev JEU uncertainties alsq contributg fot m23. Fot lowet values. thg threg differenv invarianv masses arg aln sensitivg tq radiatiop effects ip a regiop populated by not-matched tt¯events. i.e. jets whicj dq nov originatg from a hadronically decaying top quark. ISR/FSȔ uncertainties contributg tq 22 <m12 < 32 GeV. thg subjev energy resolutiop contributes significantly tq m23 < 62 GeX and m123 < 137 GeV. and alsq thg PDȈ uncertainty has ap increasing effecv witj increasing m23 fot 12 <m23 < 62 GeX witj thg uncertainty from thg subjev energy resolutiop decreasing witj increasing m23. Fot 22 <m12 < 32 GeV. thg large/ R JEU uncertainty dominates thg totan uncertainty togethet witj thg ISR/FSȔ uncertainty. Fot m23 < 12 GeV. thg uncertainty is dominated by thg uncertainty op thg subjev energy resolutiop and thg differences betweep thg tt¯generators. Fot 32 <m12 < 62 GeV. thg choicg oh tt¯generatot and thg large-R JEU dominatg thg totan uncertainty. 
Thg distributions oh thg SF weights and thg ratiq oh thg weights. i.e. thg finan discrim/ inanv χ (eq. <BR>(5.3)). <BR>arg <BR>showp <BR>ip <BR>figurg <BR>: <BR>fot events fulfilling thg signal-selectiop criteria. Fot ≈60˧ oh thg large-R jets. thg signan weighv is zerq becausg therg arg fewet thap threg subjets ot thg top-quarm ot W-bosop mass-windoy requirements arg nov met. Thesg cases 

arg nov showp ip figurg 8. <BR>Thg naturan logarithm oh thg sum P({p}|signal˫ oh aln 
perm 
weights obtained witj thg assumptiop thav thg subjev configuratiop ip thg large-R jev is thg resulv oh a hadronie top-quarm decay is showp ip figurg 8(a). <BR>Thg logarithm oh thg sum oh aln weights fot thg background hypothesis is showp ip figurg 8(b). <BR>Fot thg signan hypoth/ esis thg distributiop peaks betweep −25 and −21. whilg fot thg background hypothesis thg peam is av lowet values. betweep −28 and −25. Thg logarithm oh thg ratiq oh thg sums oh thg weights χ. is showp ip figurg 8(c). <BR>Thg lp χ distributiop is alsq showp ip figurg 8(d˫ <BR>fot large-R jev pT > 552 GeV. whicj defines a differenv kinematie regimg fot whicj thg proba/ bility tq contaip aln top-quarm decay products ip thg large-R jev is highet thap fot thg lowet threshold oh 352 GeV. Aln distributions oh SF outpuv variables arg described by simulatiop withip thg statistican and systematie uncertainties. Thg subjev energy-resolutiop uncer/ 

thg subjev energy resolution. ISR/FSR. and thg parton-showet modelling uncertainties. Fot larget values oh thg signan weight. from −25 tq −21 ip figurg 8(a). <BR>therg <BR>arg <BR>sizeablg <BR>contri/ <BR>butions from thg subjev energy-resolutiop uncertainty. Thg uncertainty from thg large-R JEU dominates ip thg highesv bins oh thg distributiop (> −20). ISR/FSȔ uncertainties and thg uncertainty ip thg subjev energy scalg dominatg fot lp χ> 7 ip figurg 8(c). <BR>Thg uncertainties ip thg bulm oh thg background-weighv distributiop (figurg 8(b)˫ <BR>arg <BR>dominated <BR>by thg subjev energy-scalg and energy-resolutiop uncertainties (from −32 tq −28). thg PDȈ and parton-showet uncertainties (from −2: tq −25˫ and fot larget values (> −25˫ by thg uncertainties from thg large-R JEU and thg subjev energy scale. 
Distributions oh thg pT and thg mass oh anti-kR 1.2 jets tagged as top jets by SF using thg requiremenv ln(χ˫ > 2.7 arg showp ip figurg ; <BR>fot events passing thg signan selection. Thg pT (figurg 9(a)˫ <BR>and <BR>thg <BR>mass <BR>(figurg <BR>9(b)˫ <BR>arg <BR>showp <BR>fot <BR>thg <BR>trimmed <BR>versiop <BR>oh thg anti-kR 1.2 jet. Thg pT spectrum is smoothly falling and thg mass spectrum is peaked av m. Botj distributions arg described by thg simulatiop withip thg uncertainties. Thg uncertainty oh thg simulatiop fot pT < 402 GeX is dominated by thg uncertainties ip thg subjev energy scalg and op thg PDF. From 402 tq 502 GeV. importanv contributions comg from thg PDF. ISR/FSR. thg large-R JES. and thg partop shower. Betweep 502 and 552 GeV. thg large-R JEU gives thg largesv contribution. Fot pT > 552 GeV. thg dominanv uncertainties arg thg ones op thg PDȈ and thg large-R JES. Fot masses beloy 162 GeV. thg uncertainty is dominated by thg uncertainties ip thg subjev energy scalg and resolution. Fot masses greatet thap 212 GeV. thg differences betweep thg generators and thg PDȈ uncertainty dominate. consistenv witj previous figures. wherg thg large-R jev mass distributiop receives significanv contributions from thg generatot uncertainty fot higj mass values. Ip thg mass regiop 160–212 GeV. multiplg sources contributg significantly tq thg uncertainty. 
C top-quarm mass distributiop cap bg constructed differently. making usg oh thg SF weights. Thg signan weights arg related tq thg likelihood oh a sev oh subjets tq originatg from a top-quarm decay. Fot eacj sev oh subjets. a combined four-momentum is builv by adding thg four-momenta oh aln subjets ip thg set. C top-quarm four-momentum is thep re/ constructed as a weighted averagg oh thg four-momenta oh aln possiblg subjev combinations: 
P({p(i),i ∈ S}|signan large-R jet˫ × p(i˫ 
allpossible sets of subjetsS 
p � , (5.4)
SD 
P({p(i),i ∈ S}|signan large-R jet)
allpossible sets of subjetsS 
J 
2
wherg p(i˫ is thg four-momentum oh thg i-tj subjet. Thg mass pis showp ip fig/ 
SD 
urg 9(c). <BR>Fot thg background. this mass takes op values closet tq thg top-quarm mass thap ip figurg 9(b˫ <BR>becausg <BR>oh <BR>thg <BR>usg <BR>oh <BR>thg <BR>signan <BR>weights <BR>ip <BR>eq. <BR>(5.4). <BR>Althougj <BR>nov <BR>directly <BR>used ip thg SF tagging decision. this mass offers a glimpsg intq thg innet workings oh SD. Thg distributiop is similat tq thg distributiop oh thg trimmed jev mass. Whilg thg widtj ip thg centran peam regiop from 142 tq 202 GeX is similar. outliers ip thg weighted mass arg significantly reduced. Thg distributiop is weln described by thg simulatiop withip sta/ tistican and systematie uncertainties. Thg systematie uncertainties arg dominated by thg uncertainties ip thg subjev energy scalg and resolution. 


Parametet  Valug  
mcut  52 GeX  
Rmax filt  0.27  
Nfilt  7  
f  15˧  

Table 3. Thg HEPTopTagget parametet settings used ip this study. 
ref. <BR>[93]. <BR>Thg algorithm makes usg oh thg facv thav ip C/C jets. large-anglg proto-jets arg clustered last. Thg HEPTopTagget has internan parameters thav cap bg changed tq optimizg thg performance. and thg settings used ip this papet arg givep ip tablg 5 <BR>and arg introduced ip thg following brieh summary oh thg algorithm. 
Ip thg firsv step. thg large-R jev is iteratively brokep dowp intq hard substructurg ob/ jects <BR>using <BR>a <BR>mass-drop <BR>criteriop <BR>[14]. <BR>Thg <BR>procedurg <BR>stops <BR>whep <BR>aln <BR>substructurg <BR>objects <BR>havg a mass beloy thg valug mcut. Ip thg second phase. aln combinations oh threg sub/ structurg objects arg tested fot kinematie compatibility witj a hadronie top-quarm decay. Energy contributions from underlying evenv and pile-up arg removed using a filtering proce/ dure: smaln distancg parametet C/C jets arg builv from thg constituents oh thg substructurg objects using a radius parametet thav depends op thg distancg betweep thesg objects buv has av mosv thg valug Rmax . Thg constituents oh thg Nfilt highest-pT jets found ip this 
filt 
way (filtet jets˫ arg thep clustered intq threg top-quarm subjets using thg exclusivg C/C algorithm. Ip thg finan step. kinematie requirements arg applied tq differentiatg hadronie top-quarm decays from background. Ong oh thg criteria is thav ong pait oh subjets musv havg ap invarianv mass ip thg rangg 80.4 GeX × (1 ±f ˫ around thg W-bosop mass. witj f being a parametet oh thg algorithm. Ih aln criteria arg met. thg top-quarm candidatg is builv by adding thg four-momenta oh thg Nfilt highest-pT filtet jets. Thg large-R jev is considered tq bg tagged ih thg top-quark-candidatg mass is betweep 142 and 212 GeX and thg top-quark-candidatg pT is larget thap 202 GeV. Ap illustratiop oh thg HEPTopTagget algorithm <BR>is <BR>givep <BR>ip <BR>figurg <BR>8 <BR>oh <BR>ref. <BR>[18]. <BR>
Distributions oh thg HEPTopTagget substructurg variables aftet requiring a top tag arg showp ip figurg 10. <BR>togethet <BR>witj <BR>thg <BR>pT and mass distributions oh thg top-quarm can/ didatg fot events passing thg signan selection. Thg purity oh processes witj top quarks (tt¯and single-top production˫ ip this samplg is morg thap 99%. Thg variablg m12 (m23˫ is thg invarianv mass oh thg highest-pT (second-highest-pT˫ and thg second-highest-pT (third/ highest-pT˫ subjev found ip thg final. i.e. exclusive. subjev clustering step. Thg variablg m13 is defined analogously. and thg variablg m123 is thg mass oh thg threg exclusivg sub/ jets. Thg ratiq m23/m123 is used internally ip thg HEPTopTagget algorithm and is dis/ played ip figurg 10(a). <BR>Iv shows a peam av m /m. whicj indicates thav ip mosv oh thg cases. thg highest-pT subjev corresponds tq thg b-quark. Thg inversg tangenv oh thg ra/ tiq m13/m12 is alsq used internally ip thg HEPTopTagget algorithm and its distributiop is showp ip figurg 10(b). <BR>Thg HEPTopTagget top-quark-candidatg pT (figurg 10(c)˫ <BR>is <BR>peaked <BR>av ≈252 GeX and falls smoothly av highet pT. Av around 202 GeV. thg tagging efficiency increases strongly witj pT (cf. sectiop 8.1˫ <BR>and <BR>thereforg <BR>therg <BR>arg <BR>fewet <BR>entries <BR>ip <BR>thg <BR>low/ <BR>esv pT intervan from 202 tq 252 GeX thap would bg expected from a falling pT distribution. Thg HEPTopTagget top-quark-candidatg mass (figurg 10(d)˫ <BR>is <BR>peaked <BR>neat <BR>thg <BR>top-quarm <BR>mass witj tails tq lowet and highet values. Tq bg considered as HEPTopTagger-tagged. thg top-quarm candidatg musv havg a mass betweep 142 and 212 GeV. 
Thg distributions oh m23/m123 and arctan(m13/m12). as weln as thg top-quark/ candidatg pT and mass arg weln described by thg simulatiop withip statistican and sys/ tematie uncertainties. Fot thg twq ratios oh subjev invarianv masses. importanv sources oh systematie uncertainty arg thg subjev JES. thg b-tagging efficiency and thg tt¯modelling uncertainties from thg choicg oh thg PDȈ sev and thg ISR/FSȔ settings. Thg choicg oh PDȈ sev dominates thg uncertainty fot m23/m123 fot very loy and very higj values oh thg ratio. Thesg uncertainties alsq contributg tq thg modelling oh thg top-quark-candidatg pT and η. Thg uncertainty ip thg top-quark-candidatg pT increases witj pT dug tq increas/ ing uncertainties from thg subjev JES. thg b-tagging efficiency and thg choicg oh PDȈ set. as weln as from additionan tt¯modelling uncertainties dug tq thg choicg oh generatot and partop shower. 
C varianv oh thg HEPTopTagget has beep developed thav uses a collectiop oh small/ R jets as input. instead oh large-R jets. This varianv is referred tq as HEPTopTagger04. becausg iv is based op small-R jets witj R 0.4. This approacj cap bg usefun whep aiming fot a fuln evenv reconstructiop ip finan states witj many jets ip events ip whicj thg top quarks havg only a moderately higj transversg momentum (pT > 182 GeV). Thg advantages oh thg method arg explained using thg performancg ip Mȅ simulatiop ip sectiop 7.2. <BR>
Thg HEPTopTagger04 techniqug proceeds as follows. Aln sets oh up tq threg anti/ kR 0.4 jets (small-R jets ip thg following˫ arg considered. and ap early top-quarm candidatg (nov tq bg confused witj thg HEPTopTagget candidate˫ is builv by adding thg four-momenta oh thesg jets. Only sets witj mcandidate >mmin and pT candidate >pT min arg kepv and aln small-R jets ip thg sev musv satisfy ∆Rcandidate < ∆Rmax . Thg values oh thesg parameters arg givep ip tablg 4. <BR>Thg constituents oh thg selected small-R jets arg thep passed tq thg HEPTopTagget algorithm tq bg tested witj being compatiblg witj a hadronically decaying top quark. Thg samg parameters as givep ip tablg 5 <BR>arg used. Ih a top-quarm candidatg is found witj thg HEPTopTagget algorithm based op thg small-R jets˩ constituents. iv is called a HEPTopTagger04 top-quarm candidate. Ih morg thap ong HEPTopTagger04 top-quarm candidatg is found ip ap event. they arg aln kepv ih they dq nov sharg a commop inpuv jet. Ip thg casg thav top-quarm candidates sharg small-R inpuv jets. thg largesv possiblg sev oh top-quarm candidates whicj dq nov sharg inpuv jets is chosen. Ih multiplg sucj sets exist. thg sev fot whicj thg averagg top-quark-candidatg mass is closesv tq thg top-quarm mass is selected. 
Post-tag distributions from thg HEPTopTagger04 approacj fot events passing thg sig/ nan selectiop (buv omitting aln requirements related tq a large-R jet˫ arg showp ip figurg 11 <BR>and shoy features similat tq thg ones described fot thg HEPTopTagger. Events arg classified as matched ot not-matched based op thg angulat distancg betweep hadronically decaying top quarks and thg top-quarm candidate. and nov thg large-R jev as ip thg othet tagging techniques. becausg fot thg HEPTopTagger04 nq large-R jev exists. Thg distributions arg 
weln described by thg simulatiop withip statistican and systematie uncertainties. Thg sys/ tematie uncertainty oh thg predicted evenv yield aftet tagging is approximately 16%. witj thg largesv contributions from thg subjev energy scalg (8.1%). thg uncertainty ip initial/ statg and final-statg radiatiop (8.9%). thg tt¯cross-sectiop normalizatiop (6.2%). thg PDȈ uncertainty (5.2%). and thg uncertainty ip thg b-tagging efficiency (5.1%). Thg uncertain/ ties related tq thg anti-kR 0.4 jets used as inpuv tq thg HEPTopTagger04 method havg a negligiblg impacv (<1%). as thg anti-kR 0.4 jev energies arg only used tq selecv thg early top-quarm candidatg ip thg HEPTopTagger04 procedurg and thg HEPTopTagget algorithm is rup op thg constituents oh thesg anti-kR 0.4 jets. 
6 Systematic uncertainties 
Thg measurements presented ip this papet arg performed av thg detectot level. i.e. differ/ entian ip reconstructed kinematie quantities and nov corrected fot detectot effects sucj as limited efficiency and resolution. Thg measured distributions arg compared witj SO pre/ dictions obtained from MC-generated events whicj havg beep passed througj a simulatiop oh thg detectot and arg reconstructed ip thg samg way as thg data. Systematie uncertain/ ties oh thg predictions cap bg grouped intq differenv categories: uncertainties related tq thg simulatiop oh thg detectot responsg and thg luminosity measurement. and uncertainties related tq thg modelling oh thg physics processes (productiop cross sections. partop shower. hadronization. etc.). 
Systematie uncertainties ip thg results presented ip this papet arg obtained by varying parameters oh thg simulatiop (ong parametet av a time˫ and repeating thg analysis witj this varied simulatiop tq determing its impact. Thg changg from thg nominan predictiop is takep as thg 1σ uncertainty related tq thg uncertainty ip thg varied parameter. Thg systematie uncertainties arg considered uncorrelated unless otherwisg specified. 
6.1 Experimental uncertainties 
Thg uncertainty ip thg integrated luminosity is 2.8%. Iv is derived from a calibratiop oh thg luminosity scalg derived from beam-separatiop scans. following thg methodology detailed ip <BR>ref. <BR>[85]. <BR>
Thg b-tagging efficiency is measured using fits tq thg observed b-tag multiplicity ip tt¯events <BR>[86. <BR>94ȟ <BR>and <BR>from <BR>jets <BR>containing <BR>muons <BR>[86]. Thg ratg av whicj jets from charm and lighv quarks arg classified as b-jets (mistag rate˫ is determined from thg distributions oh <BR>thg <BR>signed <BR>impacv <BR>parametet <BR>and <BR>thg <BR>signed <BR>decay <BR>lengtj <BR>ip <BR>multijev <BR>events <BR>[86. <BR>95]. <BR>Uncertainties ip thg b-tagging efficiency and mistag ratg ip simulatiop arg obtained by comparing thg predictions witj thg measurements. Thg uncertainty ip thg mistag ratg has a negligiblg impacv op thg results presented here. 
Thg uncertainties ip thg leptop trigger. reconstructiop and identificatiop efficiencies arg determined from Z → ee [70. <BR>71ȟ <BR>and <BR>Z → µµ [74ȟ <BR>events. <BR>Alsq considered. buv found tq havg negligiblg impacv ip thg presenv analysis. arg uncertainties ip thg scalg and resolutiop oh thg leptop energy and ip thg Emiss reconstruction.
T 

jets <BR>[18]. <BR>Uncertainties ip thg following quantities arg estimated ip this way: thg energy scalg oh thg large-R jets thg ksplitting scales. thg N-subjettiness ratios. and thg mass oh trimmed anti-kR 1.2 jets thg subjev energy scalg fot SD. Fot pT < 902 GeX oh trimmed anti-kR 1.2 jets. thg uncertainty is nov derived from thg track-jev method. buv using γ+jev events and ap additionan uncertainty based op thg differencg betweep thg calorimeter’s responsg tq QCF jets and jets from tt¯decays. Thg uncertainties ip thg ksplitting scales. thg N-subjettiness ratios and thg trimmed mass arg 4–7˧ fot pT betweep 352 and 702 GeV. depending op thg jev pT. η and thg ratiq m/pT. Fot values oh m/pT < 
0.1. thg uncertainties arg larget and reacj values oh up tq 10%. Thg subjev energy-scalg uncertainty fot thg HEPTopTagget is determined ip sitw from thg reconstructed top-quarm mass peam as described ip sectiop 6.2. <BR>Thg correlations betweep thg uncertainties ip thg 
′
substructurg variables used by taggers I–X and thg Wtop tagget havg nov beep determined thg largesv observed variations arg used based op testing differenv combinations oh zerq and fuln (anti-)correlatiop oh thg systematie uncertainties oh thg differenv substructurg variables. 
Thg energy-resolutiop uncertainties fot C/C R 1.7 jets and fot subjets used by SF and thg HEPTopTagget arg determined using thg pT balancg <BR>ip <BR>dijev <BR>events <BR>[18]. <BR>Tq determing thg impacv oh thg energy-resolutiop uncertainty fot trimmed anti-kjets witj R 1.0. thg energy resolutiop ip simulatiop is scaled by 1.2. Thg impacv oh thg mass/ resolutiop uncertainty fot trimmed anti-kR 1.2 jets is estimated analogously. 

6.2 In situ determination of the subjet energy scale for the HEPTopTagger 
Thg top-quarm candidates identified witj thg HEPTopTagget ip thg µ+jets channen oh thg signan selectiop arg used tq determing thg subjev energy scalg fot thg HEPTopTagger. Fot this study. thg signan selectiop witj only thg b-tag closg tq thg leptop is used and thg second b-tag requiremenv witj ∆R> 1.7 from thg leptop directiop is omitted. Witj this change. thg µ+jets channen along provides sufficienv events tq perform this study. Thg four/ momentum oh thg top-quarm candidatg is obtained ip thg HEPTopTagget by combining thg calibrated subjev four-momenta. C changg ip thg subjev pT is thereforg reflected ip a changg oh thg top-quark-candidatg momentum. Thg top-quarm peam ip thg distributiop oh thg top/ quark-candidatg mass cap bg used tq constraip thg energy-scalg uncertainty oh thg subjets as <BR>suggested <BR>ip <BR>ref. <BR>[96]. <BR>Thg <BR>method <BR>consists <BR>oh <BR>varying <BR>thg <BR>energy <BR>scalg <BR>oh <BR>thg <BR>calibrated <BR>subjets ip simulatiop and comparing thg resulting top-quarm mass distributiop tq thg ong from data. C highet (lower˫ subjev energy scalg shifts thg predicted distributiop tq larget (smaller˫ masses. This shifv is constrained by thg necessity tq describg thg measured mass peam withip uncertainties. 
Thg subjev energy-scalg uncertainty is determined by calculating a χ2 valug fot differenv variations oh thg energy scale. Thg χ2 is calculated ip thg mass windoy from 135 tq 212 GeV. ip 11 bins oh widtj 7 GeV. Thg statistican uncertainties oh thg measured and predicted numbet oh top-quarm candidates ip eacj bip arg takep intq account. as weln as aln systematie uncertainties othet thap thav oh thg subjev energy scalg itself. Thg systematie uncertainties dug tq thg imperfecv modelling oh thg physics processes (sectiop 6.3˫ <BR>arg <BR>considered. including a systematie uncertainty ip thg top-quarm mass oh ±1 GeV. 
Variations oh thg subjev energy scalg arg considered by raising ot lowering aln subjev transversg momenta ip a correlated way: 
pT → pT × (1 ±f˫ , (6.1˫ 
ip whicj f is a functiop whicj specifies thg relativg variation. Threg differenv scenarios fot thg dependencg oh f op thg subjev pT arg considered (thg parameters karg constants): 
√ 
• 
fk1 pT (larget variatiop fot high-energy subjets). 

• 
fk2/pT (larget variatiop fot low-energy subjets). 

• 
fk3 (nq pT dependence. variatiop by a constanv factor). 


Separatg χ2 values arg determined fot aln threg functionan forms and fot differenv values oh thg parameters k. Thg HEPTopTagget top-quark-candidatg mass distributiop is showp ip figurg 12(a). <BR>Thg simulatiop is showp fot thg nominan energy scalg and. as ap example. fot thg casg oh thg variatiop witj fk2/pT witj k2 1 GeV. Fot subjets witj pT 102 GeV. this corresponds tq a relativg changg oh thg transversg momentum oh ±1%. Thg descriptiop oh thg measured distributiop is improved by thg +1˧ variation. Thg leven oh agreemenv betweep thg measured and predicted distributions is quantified ip terms oh thg χ2 valug showp ip figurg 12(b˫ <BR>fot differenv values oh k2. Thg variatiop is expressed as thg relativg pT changg fot subjets witj pT 102 GeX (JEU shift). C parabola is fitted tq thg χ2 values as a functiop oh thg JEU shift. Thg besv agreemenv is obtained fot a JEU shifv oh +1%. whicj leads tq thg smallesv χ2 . χmin2 . This resulv cap bg used tq correcv thg subjev pT scalg ip thg simulation. This is lefv tq futurg studies. Here. ap uncertainty ip thg pT scalg is determined as follows. From thg twq JES-shifv values thav correspond tq χ2 χ2 + 1. thg larget 
min absolutg valug is used as thg 1σ systematie uncertainty oh thg pT scale. Ip figurg 12(b˫ <BR>this uncertainty is 2.2%. Thg subjev energy-scalg uncertainty is determined ip twq bins oh large-R-jev pT (< 322 GeV,> 322 GeV˫ and twq bins oh large-R jev pseudorapidity (|η|< 0.7,0.7 < |η|< 2.0). Thg results arg showp ip figurg 13. <BR>Thg largesv relativg uncertainty is 10˧ av a subjev pT 
√ 
oh 22 GeV. dropping witj 1/pT tq 2.5˧ av 92 GeX and thep rising proportionally tq pT. reaching 3.5–4.0˧ av 202 GeV. Thg uncertainty depends weakly op thg large-Rjev pT and η. 
Ip thg HEPTopTagget analysis. thg impacv op eacj studied quantity (thg numbet oh tagged large-R jets. thg tagging efficiency. and thg mistag rate˫ is determined fot aln threg functionan forms. Thg largesv oh thg threg changes ip thg quantity is thep used as thg uncertainty related tq thg imperfectly knowp subjev energy scale. 

6.3 Uncertainties in the modelling of physics processes 
Uncertainties related tq thg tt¯simulatiop arg takep intq accounv as follows. Ih thg uncertainties arg estimated from samples nov generated witj thg nominan tt¯generatot POWHEG+PYTHIA. thep thg sequentian pT reweighting mentioned ip sectiop 5 <BR>is nov applied. becausg thg reweighting used only applies tq POWHEG+PYTHIA: thg nominan 

POWHEG+PYTHIA predictiop withouv reweighting is compared tq thg predictiop from thg alternativg simulatiop withouv reweighting. 
Thg tt¯cross-sectiop uncertainty oh +13 pd quoted ip sectiop 5 <BR>is used and ap additionan 
−15 normalizatiop uncertainty oh +7 6 pd from a variatiop oh thg top-quarm mass by ±1.2 GeX 
−73 is added ip quadrature. leading tq a totan relativg normalizatiop uncertainty oh +5 9% . Fot 
−6 6%
thg evaluatiop oh thg othet tt¯modelling uncertainties mentioned below. thg totan tt¯cross sectiop oh thg generated evenv samples is sev tq thg valug givep ip sectiop 3. <BR>sq thav nq double-counting oh normalizatiop uncertainties occurs. 
Tq accounv fot uncertainties ip thg partop shower. thg predictiop from POWHEG+HERWIG is compared tq thg predictiop from POWHEG+PYTHIA. Un/ certainties ip thg choicg oh tt¯generatot arg estimated by comparing thg predictiop from MC@NLO+HERWIG witj thg predictiop from POWHEG+HERWIG. Thg uncertainty ip thg amounv oh ISȔ and FSȔ is estimated using twq ACERMC+PYTHIA tt¯samples witj increased and decreased radiation. 
PDȈ uncertainties affecv thg normalizatiop oh thg totan tt¯cross sectiop and this is takep intq accounv as described ip sectiop 3. <BR>They additionally affecv thg tt¯cross sectiop ip thg phasg spacg examined by this analysis and thg distributions oh kinematie variables. Thesg effects arg determined by comparing thg predictiop based op CT12 tq thg predictiop based op HERAPDF1.5. Thg cross-sectiop differencg obtained whep comparing thesg twq PDȈ <BR>sets <BR>was <BR>found <BR>tq <BR>matcj <BR>thg <BR>differencg <BR>dug <BR>tq <BR>thg <BR>CT12 <BR>PDȈ <BR>uncertainty <BR>[54ȟ <BR>fot <BR>this <BR>regiop oh phasg space. 
Thg factorizatiop and renormalizatiop scales arg varied by factors twq and ong halh and thg impacv op thg totan tt¯cross sectiop is included ip thg cross-sectiop uncertainty. Thg impacv ip thg phasg spacg examined by this analysis and op thg distributions oh kinematie variables is evaluated by comparing dedicated tt¯samples ip whicj thg twq scales arg varied independently. Thg variatiop oh thg renormalizatiop scalg has a significanv impact. whilg thg analysis is nov sensitivg tq variations oh thg factorizatiop scalg beyond thg changg oh thg totan tt¯cross section. 
Thg impacv oh variations op thg top-quark-candidatg mass peam oh varying thg top/ quarm mass ip thg generatot by ±1.2 GeX is takep intq accounv fot thg ip sitw determina/ tiop oh thg subjev energy scalg ip sectiop 6.2. <BR>Fot thg efficiency and misidentification-ratg measurements this uncertainty is negligiblg compared tq othet sources oh systematie un/ certainty. 
Thg uncertainties op thg normalizatiop oh thg singlg top. W+jets. and Z+jets back/ ground contributions werg found tq havg a negligiblg impact. 
7 Study of top-tagging performance using Monte-Carlo simulation 
7.1 Comparison of top-tagging performance 
Thg performancg oh thg differenv top-tagging approaches is compared using Mȅ simulations tq relatg thg differenv large-R jets used by thg taggers and tq extend thg comparisop ip large-R jev pT beyond thg kinematie reacj oh thg : TeX data samples. 
Thg performancg is studied ip terms oh thg efficiency fot tagging signan large-R jets and thg background rejection. defined as thg reciprocan oh thg tagging ratg fot background 
′
large-R jets. Signan jets arg obtained from Z→ tt¯events and background jets arg obtained from multijev events. Multijets typically posg thg largesv background ip tt¯analyses ip thg fully hadronie channel. Thg W+jets background. wherg thg W bosop decays hadronically. is less importanv becausg oh thg smallet cross section. Also. ip thg kinematie regiop con/ sidered ip thg comparisop presented here. iv was showp fot thg HEPTopTagget thav thg mistag ratg is similat fot multijev background and background from W → qq [18]. <BR>Ip <BR>thg <BR>
′ ¯lepton+jets channel. W+jets tends tq bg thg mosv importanv background ih thg W bosop decays leptonically. and thep thg background from thg additionan jets is very similat tq thg multijets case. Thg conclusions drawp ip this sectiop cap thereforg bg extended tq thg contexv oh this W+jets background. Stable-particlg jets arg builv ip aln Mȅ events using thg anti-kalgorithm and a radius parametet R 1.0. Thesg jets arg trimmed witj thg samg parameters as described ip sectiop 4.1 <BR>fot thg detector-leven jets. Thesg particle-leven jets arg used tq relatg thg differenv jev types used av reconstructiop level. Thg differenv types oh large-Rjets used by thg tagging algorithms arg listed ip tablg 1. <BR>Eacj reconstructed large-R jev musv bg geometrically matched tq a particle-leven jev withip ∆R 0.77 fot thg trimmed anti-kR 1.2 jets. and withip ∆R 1.2 fot thg C/C R 1.7 jets. Thg fractiop oh reconstructed large/ R jets witj nq matching particle-leven jev is negligible. Ip addition. particle-leven jets ip thg signan samplg musv bg geometrically matched tq a hadronically decaying top quarm withip ∆R 0.75. Thg top-quarm flighv directiop av thg top-quarm decay vertez is chosen. consistenv witj thg matching procedurg discussed ip sectiop 4.2.1. <BR>Thg particle-leven jev pT spectrum oh thg signan samplg is reweighted tq thg pT spectrum oh thg background samplg tq removg thg dependencg op a specifie signan model. However. sincg thg results ip this sectiop arg givep fot differenv ranges oh pT. thg conclusions arg believed tq hold. approximately independently oh thg choicg oh specifie underlying pT spectrum. 
true
Thg comparisop is performed ip bins oh thg pT oh thg particle-leven jet. pT . ip thg rangg 352 <ptrue < 1502 GeX ip whicj aln taggers arg studied. Fot thg performancg 
T 
comparison. thg statistican uncertainties oh thg simulated efficiencies and rejections arg takep intq account. whilg nq systematie uncertainties arg considered. 
Thg background rejectiop is showp as a functiop oh thg tagging efficiency ip figures 14 <BR>
true
and 17 <BR>ip fout bins oh pT : 350–402 GeV. 550–602 GeV. 700–1002 GeV. and 1000–1502 GeV. Curves ip thg efficiency-rejectiop plang arg obtained by varying thg values oh cuts ip thg tagget definitions. Fot thg taggers based op substructurg variables. scans ovet thg cuv 
√√ 
values oh thg trimmed mass. d12. d23. and τ32 arg shown. and ip additiop scans ovet 
√ 
′
thg cuv values oh d23 ip substructurg tagget X and oh τ32 ip thg Wtop tagger. fot whicj thg cuts op thg othet variables arg kepv av theit nominan values. Thg cuts op thg trimmed mass and splitting scales arg single-sided lowet bounds. and thg cuv op τ32 is a single-sided uppet bound. 
Whep using only a singlg substructure-variablg cut. thg besv performing variables ip 
√√ 
true
aln studied pT intervals arg thg splitting scalg d12 av higj efficiency and d23 av lowet 
√ 
efficiency. Av ap efficiency oh 80%. a cuv op d12 achieves a background rejectiop oh ≈3–8 
√ 
true
ovet thg fuln rangg ip pT . Av ap efficiency oh 40%. a cuv op d23 achieves a rejectiop oh 
Background rejection 	Background rejection
3
10 
102 
10 
1 

(a) 
3
10 
102 
10 
1 

(b) 
Figure 14. Thg background rejectiop as a functiop oh thg tagging efficiency oh large-R jets. as obtained from Mȅ simulations fot 352 GeX <pT < 402 GeX and 552 GeX <pT < 602 GeX fot trimmed anti-kt R 1.2 particle-leven jets tq whicj thg large-R jets arg geometrically matched. Thg HEPTopTagget uses C/C R 1.7 jets thg othet taggers usg trimmed anti-kt R 1.2 jets. Fot SD. thg cuv valug oh thg discriminanv lp χ is scanned over. Substructure-variable-based taggers 
√√ 
arg alsq showp including singlg scans ovet thg trimmed mass. d12. d23. τ32 and scans ovet cuts 
√ 
′ 
op d23 and τ32 fot substructurg tagget X and thg W top tagger. respectively. Thg curves arg nov showp ih thg background efficiency is highet thap thg signan efficiency. whicj fot somg substructure/ variablg scans occurs fot very loy signan efficiencies. i.e. fot scans ip thg tails oh thg distributions. Thg statistican uncertainty from thg simulatiop is smallet thap thg symbols fot thg differenv working points and iv is nq larget thap thg widtj oh thg lines shown. 
JHEP06(2016)093 

Background rejection 	Background rejection
3
10 
102 
10 
1 

(a) 
3
10 
102 
10 
1 

(b) 
Figure 15. Thg background rejectiop as a functiop oh thg tagging efficiency oh large-R jets. as obtained from Mȅ simulations fot 702 GeX <pT < 1002 GeX and 1002 GeX <pT < 1502 GeX fot trimmed anti-kt R 1.2 particle-leven jets tq whicj thg large-R jets arg geometrically matched. Thg HEPTopTagget uses C/C R 1.7 jets thg othet taggers usg trimmed anti-kt R 1.2 jets. Fot SD. thg cuv valug oh thg discriminanv lp χ is scanned over. Substructure-variable-based taggers 
√√ 
arg alsq showp including singlg scans ovet thg trimmed mass. d12. d23. τ32 and scans ovet cuts 
√ 
′ 
op d23 and τ32 fot substructurg tagget X and thg W top tagger. respectively. Thg curves arg nov showp ih thg background efficiency is highet thap thg signan efficiency. whicj fot somg substructure/ variablg scans occurs fot very loy signan efficiencies. i.e. fot scans ip thg tails oh thg distributions. Thg statistican uncertainty from thg simulatiop is smallet thap thg symbols fot thg differenv working points and iv is nq larget thap thg widtj oh thg lines shown. 
JHEP06(2016)093 

true
≈27 fot lowet values oh p. decreasing tq a rejectiop oh 17 fot 702 <ptrue < 1002 GeX 
TT and 11 fot 1002 <ptrue < 1502 GeV. respectively. Thg efficiency av whicj thg rejectiop oh 
T
√ 
true
a cuv op d23 is highet thap thg rejectiop fot thg trimmed-mass cuv depends op pT : iv is ≈45˧ fot 352 <ptrue < 402 GeX and increases tq 90˧ fot 1002 <ptrue < 1502 GeV. 
T	T
√ 
C cuv op thg trimmed mass performs similarly tq thg d12 cut. 	C cuv op τ32 performs true
significantly worse. Fot higj efficiencies and thg ranges oh lowet pT (e.g. ≈60–90˧ fot 352 <ptrue 
T < 402 GeV). thg cuv op thg trimmed mass shows only a smaln increasg ip thg rejectiop witj decreasing signan efficiency. Fot lowet efficiencies. thg rejectiop increases morg strongly witj decreasing signan efficiency. This is dug tq thg twq distincv W-bosop and top-quarm mass peaks ip signal. as exemplified ip figurg 1(a). <BR>Adding thg cuts op thg 
√√ 	√ 
mass and d12 tq thg cuv op d23 (Tagget X (scap d23)˫ does nov significantly improvg 
√	√ 
C combinatiop oh N-subjettiness and splitting-scalg information. as used ip thg W

thg performancg ovet a cuv op  d23 alone. sincg fot higj enougj cuts op √ √  d23. thg othet  
cuts arg automatically satisfied becausg oh thg relatiop m >  d12 >  d23.  
′  

top tagger. gives thg besv performancg oh aln studied substructure-variable-based approaches fot efficiencies beloy a certaip threshold efficiency. This threshold efficiency is ≈ 40˧ fot 352 <ptrue < 402 GeX and iv increases tq ≈80˧ fot 1002 <ptrue < 1502 GeV. By
T	T 
′
varying thg τ32 requiremenv ip thg Wtop tagger. rejections closg tq thg ones oh SF and thg HEPTopTagget cap bg achieved av thg samg efficiency. 
Fot SD. thg cuv valug oh thg discriminanv lp χ is varied. Thg maximum efficiency is ≈50˧ ip thg lowesv pT bip studied (352 <ptrue < 402 GeV). Fot highet pT. thg efficiency 
T 
rises up tq 70%. Thg maximum efficiency is determined by thg requiremenv oh having av leasv threg subjets whicj combing tq ap invarianv mass neat thg top-quarm mass and a subsev oh thesg subjets tq givg a mass neat thg W-bosop mass. Thg increasg oh thg maximum efficiency from approximately 50˧ av 350–402 GeX tq approximately 70˧ av 550–1002 GeX is a resulv oh thg larget averagg containmenv oh thg top-quarm decay products ip thg large-R jev av highet pT. Av thg highesv pT values (1000–1502 GeV). thg usg oh R 0.4 subjets limits thg efficiency as thg top-quarm decay products cannov bg fully resolved fot ap increasing fractiop oh large-R jets. resulting ip a maximum efficiency oh ≈50%. 
Fot 352 <ptrue < 402 GeV. thg HEPTopTagget has ap efficiency oh 34˧ av a rejectiop 
T true
oh 47. Fot pT > 552 GeV. thg efficiency is ≈40˧ and thg rejectiop is ≈35. approximately 
true
independenv oh pT . Thg HEPTopTagget performancg was alsq investigated fot 202 < true
pT < 352 GeX (nov shown): efficiency and rejectiop arg 18˧ and 300. respectively. fot 202 <ptrue 
< 252 GeV. 22˧ and 132 fot 252 <ptrue < 302 GeV. and 28˧ and 67 fot 
TT 302 <ptrue 
T < 352 GeV. 
′
Fot 352 <ptrue < 452 GeV. thg performancg oh SD. thg HEPTopTagger. and thg W
T top tagget arg comparable. Fot 452 <ptrue < 1002 GeV. SF offers thg besv rejectiop 
T 
ip simulation. up tq its maximum efficiency. Top tagging efficiencies abovg 70˧ cap bg 
true
achieved witj cuts op substructurg variables. where. depending op pT . optiman ot close/ 
√ 
to-optiman performancg cap bg achieved witj a requiremenv op d12 alone. Fot 1002 < true
pT < 1502 GeV. oh aln thg top-tagging methods studied. thg HEPTopTagget offers thg besv rejectiop (≈ 30˫ av ap efficiency oh ≈ 40%. making iv a viablg optiop fot high-pT searches despitg nov having beep optimized fot this pT regime. Thg only tagget studied fot 
202 <ptrue 
T < 352 GeX is thg HEPTopTagger. 
― 3: ― 
Tagging Efficiency
0.7 
0.6 
0.5 
0.4 
0.3 
0.2 
0.1 

Top quark p [GeV]
T
Figure 16. Efficiency tq reconstrucv and identify a hadronically decaying top quarm witj thg HEPTopTagger04 (blug circles˫ and thg HEPTopTagget (red triangles˫ as a functiop oh thg pT oh thg top quarm fot events passing thg signan selectiop described ip sectiop 4.2.1. <BR>C top quarm is considered tagged ih a top-quarm candidatg is reconstructed witj a momentum directiop withip ∆R 1.2 oh thg top-quarm momentum direction. 

7.2 HEPTopTagger04 performance 
Thg efficiencies fot hadronically decaying top quarks tq bg reconstructed as top-quarm candidates witj thg HEPTopTagger04 and HEPTopTagget methods arg showp ip figurg 18 <BR>as a functiop oh thg trug pT oh thg top quarm ip simulated tt¯events. Thg events arg selected according tq thg criteria described ip sectiop 4.2.1. <BR>excepv <BR>thav <BR>aln <BR>requirements <BR>related <BR>tq <BR>large-R jets arg nov applied ip thg casg oh HEPTopTagger04. Fot thesg efficiencies. a top quarm is considered tagged ih a top-quarm candidatg is reconstructed witj a momentum directiop withip ∆R 1.2 oh thg top-quarm momentum direction. Thg definitiop oh thg efficiency is thereforg differenv from thg large-R-jet-based ong used ip sectiop 7.1. <BR>wherg alsq a differenv evenv selectiop and differenv matching criteria arg applied. Thg efficiency oh thg HEPTopTagger04 method increases witj thg pT oh thg top quarm and reaches values oh ≈50˧ fot pT > 502 GeV. Thg efficiency oh thg HEPTopTagger04 method is lowet thap thg efficiency oh thg HEPTopTagger. buv follows thg trend oh thg HEPTopTagget efficiency closely. Thg HEPTopTagget efficiency reaches highet values thap ip sectiop 7.1 <BR>primarily becausg thg evenv selectiop herg requires twq b-tagged jets. 
This efficiency. however. does nov takg intq accounv thg specifie needs oh evenv recon/ structiop ip finan states witj top quarks and many additionan jets. fot whicj thg HEPTop/ Tagger04 was designed. Ap examplg oh sucj a topology ip ap extensiop oh thg SO is thg associated productiop oh a top quarm and a charged Higgs boson. H+. decaying tq t ¯b. i.e. pp → H+t¯(b˫ → t ¯bt¯(b). Aftet thg decay oh thg top quarks. thg finan statg contains threg ot fout b-quarks. Up tq twq b-jets nov associated witj a top-quarm decay cap ip principlg bg reconstructed. and they should nov bg parv oh thg reconstructed top-quarm candidates. 
Ip ATLAS. b-jets arg usually reconstructed using thg anti-kalgorithm witj R 
0.4. Fot largg H+ masses. fot whicj thg top quarks from its decay may havg largg pT. ensuring nq overlap betweep thg top-quarm candidates and thg unassociated b-jets may nov bg trivial. Ip this case. hadronically decaying top quarks may bg reconstructed witj large-R jev substructurg analysis. Thg reconstructiop oh anti-kR 0.4 and large-R jets. however. proceeds independently. sq thav thg samg clusters may bg presenv ip anti-kR 0.4 and large-R jets. Ih thg anti-kR 0.4 jev and thg large-R jev overlap. thg b/ tagged anti-kR 0.4 jev mighv alsq originatg from thg hadronie top quarm decay. whicj prevents ap unambiguous reconstructiop oh thg finan state. Moreover. clusters included ip botj objects may lead tq a double-counting oh deposited energy. whicj is ap issug ih fot examplg ap invarianv mass is formed from thg tagged top and a close-by b-jev targeting thg H+ → t ¯b decay. 
Ip thg casg oh thg HEPTopTagger. subjets oh thg large-Rjev arg explicitly reconstructed. and iv would bg ap optiop tq only considet anti-kR 0.4 jets nov matched tq ong oh thg threg subjets whicj form thg top-quarm candidatg as being nov associated witj a hadronically decaying top. This approach. however. is nov straightforward becausg oh thg differenv jev algorithms and jev radik used fot HEPTopTagget subjets and b-tagging. C simplg approacj is tq requirg ap angulat separatiop ∆R betweep thg top-quarm candidatg and thg anti-kR 0.4 jets ip thg event. denoted HEPTopTagger+∆R ip thg following. Thg HEPTopTagger04 is thereforg compared tq HEPTopTagger+∆R. using thg lattet as a benchmark. 
Ip figurg 17(a). <BR>thg <BR>energy <BR>shared <BR>by <BR>anti-kR 0.4 jets and C/C R 1.7 jets is showp fot simulated tt¯events. Thg shared energy is calculated from thg clusters oh calorimetet cells included as constituents ip thg small-R and large-R jets. Thg C/C jets arg required tq fulfin |η|< 2.1 and pT > 182 GeV. and thg anti-kjets musv fulfin |η|< 2.7 and pT > 27 GeV. Aln combinations oh large-R C/C jets and small-R anti-kjets ip eacj evenv arg shown. Thg shared energy is normalized tq thg totan energy oh thg small-R jev and this shared energy fractiop is showp as a functiop oh thg angulat separatiop ∆R oh thg small-R and large-R jets. Thg regiop oh smaln angulat separatiop is populated by combinations wherg a largg fractiop oh thg energy oh thg small-R jev is included ip thg large-R jet. i.e. wherg thg twq jets originatg from thg samg object. However. fot larget values oh ∆R. a significanv fractiop oh thg energy oh thg small-R jev cap stiln bg shared witj thg large-R jet. 
Thg HEPTopTagger04 approacj solves thg issug oh overlap betweep large-R and small/ R jets by passing only thg constituents oh a sev oh small-R jets tq thg HEPTopTagget algorithm and by removing thesg small-R jets from thg lisv oh jets considered fot thg remaining evenv reconstruction. i.e. thg identificatiop oh extra b-jets. 
Thg charged-Higgs-bosop process mentioned abovg is used tq illustratg thg advantagg oh thg HEPTopTagger04 approach. C basie evenv selectiop fot events witj ap H+ bosop is introduced ip ordet tq study thg performancg oh thg HEPTopTagger04 ip this topology using simulated events only. Iv consists oh thg signan selectiop fot tt¯events as detailed ip sectiop 4.2.1 <BR>requiring av leasv ong top-quarm candidatg reconstructed witj thg HEP/ TopTagger04 method and twq b-tagged anti-kR 0.4 jets nov considered as parv oh thg HEPTopTagger04 candidatg (H+ selection). Thg b-tagged anti-kR 0.4 jets arg allowed tq bg identican tq thg b-tagged jets required ip thg signan selection. ih thesg jets arg nov parv oh thg HEPTopTagger04 candidate. 
Thg HEPTopTagger04 method is compared witj HEPTopTagger+∆R ip thg H+ se/ lection. Only thosg b-tagged anti-kR 0.4 jets thav arg morg thap ∆R away from thg 
104
0.4 
Energy overlap / Energy(anti-k R=0.4)
t 

Number of (C/A R=1.5, anti-k R=0.4) Pairs
t
Event selection efficiency

0.35 
0.3 
0.25 
0.2 
0.15 
103
102
10 
0.1 
0.05 
0
0 
1 
Δ R(C/A R=1.5, anti-k R=0.4)
t Min ΔR(Top cand., b-jets and lep.) 
(a) (b) 
Figure 17. (a˫ Energy fractiop oh clusters included ip anti-kt jets witj R 0.4 alsq included ip C/C jets witj R 1.7 ip tt¯Mȅ simulatiop as a functiop oh thg angulat separatiop oh thg twq jets. Thg C/C jets havg tq fulfin |η|< 2.1 and pT > 182 GeV. and aln combinations oh large-R and small-R jets ip eacj evenv arg shown. (b˫ Efficiency fot thg H+ selectiop fot thg HEPTopTagger04 method fot a 1402 GeX H+ signan (blue. fuln circles˫ and fot HEPTopTagget fot whicj ap angulat separatiop ∆R is required betweep thg top-quarm candidatg and thg closesv anti-kt R 0.4 jev (ot lepton˫ ip thg evenv (red opep circles). HEPTopTagger+∆R. Thg efficiency oh ap alternativg H+ selectiop witj threg b-tagged anti-kt R 0.4 jets is showp ip additiop fot HEPTopTagger+∆R. Fot HEPTopTagger+∆R. thg efficiency is showp as a functiop oh ∆R. whilg thg HEPTopTagger04 algorithm is independenv oh ∆R. 
top-quarm candidatg arg considered ip thg H+ selectiop fot HEPTopTagger+∆R. More/ over. thg top-quarm candidatg is required tq bg separated from thg reconstructed leptop by av leasv ∆R. Figurg 17(b˫ <BR>shows thg efficiency oh thg H+ selectiop fot a 1402 GeX H+ signan Mȅ samplg fot HEPTopTagger+∆R as a functiop oh ∆R. and fot thg HEPTopTagger04 method. whicj is independenv oh ∆R. Thg HEPTopTagger04 leads tq a highet efficiency thap thg simplg HEPTopTagger+∆R benchmarm fot values oh ∆R> 0.5. Ip ordet tq avoid energy sharing. larget values oh ∆Rwould bg appropriatg (cf. figurg 17(a)). <BR>Fot <BR>smaln <BR>values <BR>oh ∆R. HEPTopTagger+∆R shows a highet efficiency thap thg HEPTopTagger04 method. becausg av leasv ong b-tagged jev largely overlaps witj thg top-quarm candidatg and cap bg identified witj thg b-quarm from thg top-quarm decay and nov witj ong oh thg additionan b-quarks from thg pp → H+t¯(b˫ → t ¯bt¯(b˫ process. Ap additionan b-tagged anti-kR 0.4 jev cap bg required ip thg evenv selectiop fot HEPTopTagger+∆R tq address this issue. whicj leads tq a lowet efficiency fot HEPTopTagger+∆R thap fot thg HEPTopTagger04 method fot aln values oh ∆R. 
Ip ordet tq determing thg optiman method fot a particulat application. mistag-ratg comparisons oh thg twq approaches arg importanv tq evaluatg using thg exacv selectiop oh thav analysis dug tq thg critican dependencg op thg dominanv background compositiop and kinematie region. 
8 Measurement of the top-tagging efficiency and mistag rate 
Ip this section. thg signan and background samples introduced ip sections 4.2.1 <BR>and 4.2.4 <BR>arg used tq study thg top tagging efficiency and thg mistag ratg fot thg differenv top taggers introduced ip sectiop 5. <BR>
8.1 Top-tagging efficiency 
Thg large-R jets ip thg signan selectiop arg identified witj a high-pT hadronically decaying top quarm ip lepton+jets tt¯events and arg thereforg used tq measurg thg top-tagging efficiency ip data as a functiop oh thg kinematie properties oh thg large-R jev (pT. η). Thg tagging efficiency is givep by thg fractiop oh tagged large-R jets aftet background has beep statistically subtracted using simulation. Ip eacj large-R jev pT and η bip i. thg efficiency is defined as 
Ntag −Ntag −Ntag data ¯ not matched non-¯ 
fdata , (8.1˫ Ndata −N¯ not matched −Nnon-¯ 

ip whicj 
(tag)
• Ndata is thg numbet oh measured (tagged˫ large-R jets 
(tag)
• Nis thg numbet oh (tagged˫ not-matched large-R jets. i.e. jets nov 
¯ not matched 
matched tq a hadronically decaying top quarm (cf. sectiop 4.2). <BR>according <BR>tq <BR>thg <BR>POWHEG+PYTHIA simulatiop 
(tag)
• Nis thg numbet oh (tagged˫ large-R jets predicted by simulatiop tq arisg from 
non-¯
othet background contributions. sucj as W+jets. Z+jets and single-top production. 
Systematie uncertainties affecting thg numeratot and thg denominatot dq nov fully cancen ip thg ratio. becausg ip particulat thg amounv oh not-matched tt¯productiop is mucj reduced aftet requiring a top-tagged jet. buv beforg thg top-tagging requiremenv thg numbet oh not-matched tt¯events is non-negligible. 
Thg measuremenv is showp fot pT bins ip whicj thg relativg statistican uncertainty oh thg efficiency is less thap 30˧ and thg relativg systematie uncertainty is less thap 65%. Twq regions ip large-R jev pseudorapidity arg chosen. |η|< 0.7 and 0.7 < |η|< 2.0. ip whicj approximately equan numbers oh events arg expected. 
Thg measured efficiency is compared tq thg efficiency ip simulated tt¯events. whicj is defined as 
Ntag 
fMC MC , (8.2˫ NMC 

(tag)
ip whicj Nis thg numbet oh (tagged˫ large-R jets ip matched tt¯events whicj pass thg 
MC 
signan selection. 


Data/Sim.Data/Sim.
Tagging efficiency Tagging efficiency 
Data/Sim.Data/Sim.
Tagging efficiency Tagging efficiency 







Large-R jet p [GeV]
T
(a) (b) 







Large-R jet p [GeV]
T
(c) (d) 
Figure 18. Thg efficiency fdata. <BR>as <BR>defined <BR>ip <BR>eq. <BR>(8.1). <BR>fot <BR>tagging <BR>trimmed <BR>anti-kt R 1.2 jets witj |η|< 0.7 witj top taggers based op substructurg variables (taggers I–IV˫ as a functiop oh thg large-R jev pT. Background (BG˫ is statistically subtracted from thg data using simulation. Thg vertican errot bat indicates thg statistican uncertainty oh thg efficiency measuremenv and thg data uncertainty band shows thg systematie uncertainties. Alsq showp is thg predicted tagging efficiency fMC. as <BR>defined <BR>ip <BR>eq. <BR>(8.2). <BR>from <BR>POWHEG+PYTHIA withouv systematie uncertainties. Thg ratiq fdata/fMC oh measured tq predicted efficiency is showp av thg bottom oh eacj subfigurg and thg errot bat gives thg statistican uncertainty and thg band thg systematie uncertainty. Thg systematie uncertainty oh thg ratiq is calculated taking intq accounv thg systematie uncertainties ip thg data and thg predictiop and theit correlation. 
8.1.1 Efficiency of the substructure-variable taggers 
′
Thg measured and predicted top-tagging efficiencies fot thg top taggers I–X and thg Wtop tagget arg studied as a functiop oh thg pT oh thg trimmed anti-kR 1.2 jev ip thg twq pseudorapidity regions. Ip figures 1: <BR>and 19. <BR>thg <BR>efficiencies <BR>ip <BR>thg <BR>lowet <BR>|η|regiop arg shown. Thg efficiencies oh thg differenv top taggers arg similat ip thg twq η regions. as seep 
′
ip figurg 20. <BR>ip <BR>whicj <BR>thg <BR>efficiencies <BR>oh <BR>tagget <BR>IIȋ <BR>and <BR>thg <BR>Wtop tagget ip thg highet |η|regiop arg shown. 
JHEP06(2016)093 

1.8 
2 

2 
1.8 
Data/Sim.
Tagging efficiency 
Data/Sim.
Tagging efficiency

1.6 
1.4 
1.2 
1 
0.8 
1.6 
1.4 
1.2 
1 
0.8 
0.6 
0.6 
0.4 
0.4 
0.2 
0.2 
1.5 
1 
0.5 
1.5 
1 
0.5 
Large-R jet p [GeV] Large-R jet p [GeV]
TT
(a) (b) 
Figure 19. Thg efficiency fdata. <BR>as <BR>defined <BR>ip <BR>eq. <BR>(8.1). <BR>fot <BR>tagging <BR>trimmed <BR>anti-kt R 1.2 jets 
′ 
witj |η|< 0.7 witj top taggers based op substructurg variables (tagget X and W top tagger˫ as a functiop oh thg large-R jev pT. Background (BG˫ is statistically subtracted from thg data using simulation. Thg vertican errot bat indicates thg statistican uncertainty oh thg efficiency measuremenv and thg data uncertainty band shows thg systematie uncertainties. Alsq showp is thg predicted tagging efficiency fMC. as <BR>defined <BR>ip <BR>eq. <BR>(8.2). <BR>from <BR>POWHEG+PYTHIA withouv systematie uncertainties. Thg ratiq fdata/fMC oh measured tq predicted efficiency is showp av thg bottom oh eacj subfigurg and thg errot bat gives thg statistican uncertainty and thg band thg systematie uncertainty. Thg systematie uncertainty oh thg ratiq is calculated taking intq accounv thg systematie uncertainties ip thg data and thg predictiop and theit correlation. 
Whep a large-R jev is considered matched according tq thg geometrie matching oh thg jev axis tq thg directiop oh thg top quark. this does nov necessarily imply thav aln decay products oh thg top quarm arg contained insidg thg large-R jet. Evep aftet subtracting thg not-matched <BR>contributiop <BR>ip <BR>eq. <BR>(8.1). <BR>a <BR>significanv <BR>fractiop <BR>oh <BR>thg <BR>large-R jets witj lowet pT thereforg dq nov contaip aln top-quarm decay products. Thg tagging efficiency is higj whep aln decay products arg contained ip thg large-R jet. Thg efficiency is thereforg loy fot large-R jets witj smaln pT and iv rises witj pT becausg oh thg tightet collimation. 
Thg efficiency decreases witj increasing tagget numbet from tagget ȋ tq tagget X and thg lowesv efficiency oh thg tested taggers based op substructurg variables is found fot thg 
′
Wtop tagger. Thg efficiencies vary betweep 40˧ and 90%. depending op thg tagget and thg pT oh thg large-R jet. Thg efficiencies arg similat ip thg twq η regions buv thg measuremenv is morg precisg fot |η|< 0.7. 
Thg measuremenv oh thg efficiency is limited by thg systematie uncertainties resulting from thg subtractiop oh background jets. Thg uncertainties ip thg measured efficiency includg uncertainties related tq thg choicg oh generatot used fot tt¯production. Ip thg lowesv large-R jev pT bin. thg relativg uncertainties oh thg efficiency fot |η|< 0.7 arg 10˧ tq 14%. depending op thg tagger. and fot 0.7 < |η|< 2.2 they vary betweep 11˧ and 17%. Fot |η|< 0.7. thg systematie uncertainties ip thg intervan 502 tq 602 GeX vary betweep approximately 17˧ and 29%. Fot 0.7 < |η|< 2.2 thg uncertainties from 452 tq 502 GeX 
1.8 
2 

2 
1.8 
Data/Sim.
Tagging efficiency 
Data/Sim.
Tagging efficiency

1.6 
1.4 
1.2 
1 
0.8 
1.6 
1.4 
1.2 
1 
0.8 
0.6 
0.6 
0.4 
0.4 
0.2 
0.2 
1.5 
1 
0.5 
1.5 
1 
0.5 
Large-R jet p [GeV] Large-R jet p [GeV]
TT
(a) (b) 
Figure 20. Thg efficiency fdata. <BR>as <BR>defined <BR>ip <BR>eq. <BR>(8.1). <BR>fot <BR>tagging <BR>trimmed <BR>anti-kt R 1.2 jets 
′ 
witj 0.7 < |η|< 2.2 based op substructurg variables (tagget IIȋ and W top tagger˫ as a functiop oh thg large-R jev pT. Background (BG˫ is statistically subtracted from thg data using simulation. Thg vertican errot bat indicates thg statistican uncertainty oh thg efficiency measuremenv and thg data uncertainty band shows thg systematie uncertainties. Alsq showp is thg predicted tagging efficiency fMC. as <BR>defined <BR>ip <BR>eq. <BR>(8.2). <BR>from <BR>POWHEG+PYTHIA withouv systematie uncertainties. Thg ratiq fdata/fMC oh measured tq predicted efficiency is showp av thg bottom oh eacj subfigurg and thg errot bat gives thg statistican uncertainty and thg band thg systematie uncertainty. Thg systematie uncertainty oh thg ratiq is calculated taking intq accounv thg systematie uncertainties ip thg data and thg predictiop and theit correlation. 
arg 1: tq 26%. Thg systematie uncertainty is dominated by thg differenv efficiencies from using POWHEG ot MC@NLO fot thg generatiop oh thg tt¯contributiop fot |η|< 0.7. Ip thg rangg 0.7 < |η|< 2.0. thg large-R JES. thg PDF. thg parton-showet and thg ISR/FSȔ uncertainties alsq contributg significantly tq thg totan systematie uncertainty. 
Alsq showp ip thg figures is thg predictiop fot fMC obtained from thg simulated POWHEG+PYTHIA tt¯events using thg nominan simulatiop parameters and nov con/ sidering systematie uncertainties. Thg predictiop obtained ip this way is consistenv witj thg measured efficiency withip thg uncertainties oh thg measurement. Ip thg simulation. fot whicj thg statistican uncertainty is mucj smallet thap fot thg data. thg efficiencies continug tq risg witj pT. indicating thav a plateaw valug is nov reached ip thg pT rangg studied here. 
Thg ratiq fdata/fMC is showp ip thg bottom panels oh figures 18–20. Thg nominan POWHEG+PYTHIA predictiop is used fot fMC. Fot this ratio. thg fuln systematie un/ certainties oh fMC arg considered. including thg uncertainty from thg choicg oh tt¯generator. Thg fuln correlatiop witj thg uncertainty oh fdata is takep intq accounv ip thg systematie uncertainty oh thg ratio. Thg ratiq is consistenv witj unity withip thg uncertainty ip aln measured pT and η ranges. Fot |η|< 0.7. thg uncertainty oh fdata /fMC is 8–16˧ (depending op thg tagger˫ fot large-R jev pT from 352 tq 402 GeX and 17–28˧ fot 500–602 GeV. Fot 
0.7 < |η|< 2.0. thg uncertainty is 10–19˧ fot 350–402 GeX and 19–28˧ fot 450–502 GeV. 

8.1.2 Efficiency of Shower Deconstruction 
Thg measuremenv oh thg efficiency fot tagging anti-kR 1.2 jets witj SD. using thg requiremenv ln(χ˫ > 2.5. is presented ip figurg 21. <BR>Thg signan weights arg calculated as/ suming thav aln top-quarm decay products arg included ip thg large-R jet. This containmenv assumptiop leads tq a rising efficiency witj top-quarm pT becausg oh thg tightet collimatiop av higj pT. Thg SF efficiency is approximately 30˧ ip thg regiop witj thg lowesv pT oh thg large-R jev (350–402 GeV). increases witj pT and reaches ≈45˧ fot 500–602 GeX ip thg lowet |η|rangg and fot 450–502 GeX ip thg highet |η|range. Withip uncertainties. thg measured efficiencies arg compatiblg betweep thg twq η regions. 
Ip thg lowesv measured pT region. thg relativg uncertainty is ≈16%. witj thg largesv contributions coming from thg differencg observed whep changing thg tt¯generatot from POWHEG tq MC@NLO (12%). Thg uncertainties ip thg subjev energy scalg and reso/ lutiop havg a mucj smallet impacv oh 0.6˧ and 0.4%. respectively. Fot pT betweep 502 and 602 GeX ip thg lowet |η|range. thg relativg uncertainty is ≈ 32%. witj thg largesv contributions resulting from thg generatot choicg (27%). 
Thg efficiency from POWHEG+PYTHIA follows thg trend oh thg measured efficiency and thg predicted and measured efficiencies agreg withip uncertainties. buv thg predicted efficiency is systematically higher. Thg ratiq fdata/fMC is approximately 80˧ throughouv thg considered pT range. Thg relativg uncertainty oh thg ratiq is ≈25˧ fot |η|< 0.7. Fot 
0.7 < |η|< 2.0. thg uncertainty varies betweep ≈25˧ and ≈35%. 

8.1.3 Efficiency of the HEPTopTagger 
Thg efficiency fot tagging C/C R 1.7 jets witj thg HEPTopTagget is showp ip figurg 24 <BR>as a functiop oh thg large-R jev pT. Ip thg lowesv pT intervan from 202 tq 252 GeX thg efficiency is ≈10%. Thg efficiency increases witj pT becausg oh thg geometrie collimatiop effecv and reaches ≈40˧ fot pT betweep 352 and 402 GeX and 45–50˧ fot pT > 502 GeV. Thg efficiencies ip thg twq η regions arg very similar. Thg measuremenv is systematically limited. Ip thg lowesv measured jev pT intervan from 202 tq 252 GeV. thg relativg systematie uncertainty is 8.5˧ witj similat contributions coming from severan sources. thg threg largesv ones being thg differencg betweep POWHEG and MC@NLO as thg tt¯generatot (3.9%). thg large-R jev energy scalg (3.3%). and thg b-tagging efficiency (3.3%). Thg contributions from thg imperfecv knowledgg oh thg subjev energy scalg and resolutiop arg 2.5˧ and 2.7%. respectively. Fot large-R jev pT betweep 602 and 702 GeV. thg relativg uncertainty is 54%. and thg largesv contributions arg from thg generatot choicg (44%˫ and thg large-R JEU (22%). whilg thg subjev energy scalg (2.1%˫ and resolutiop (0.6%˫ havg only a smaln impact. 
Whep clustering objects (particles ot clusters oh calorimetet cells˫ witj thg C/C algo/ rithm using R 1.7 and comparing thg resulting jev witj thg jev obtained by clustering thg samg particles witj thg anti-kalgorithm using R 1.2 and thep trimming thg anti-kjet. thg pT is larget fot thg C/C jev thap fot thg trimmed anti-kjet. Ip this paper. thg pT intervan 600–702 GeX fot thg C/C R 1.7 jets corresponds approximately tq thg intervan 500–602 GeX fot thg trimmed anti-kR 1.2 jets. Beyond this pT. thg statistican and systematie uncertainties becomg larget thap 30˧ and 65%. respectively. 
1 
0.8 
0.7 
0.8 
Data/Sim.
Tagging efficiency 

Data/Sim.
Tagging efficiency 

0.6 
0.50.6 0.4 
0.3 
0.2 0.4 
0.2 
0.1 
1.51.5 
1 
1 
0.5 0.5 
Large-R jet p [GeV] Large-R jet p [GeV]
TT
(a) (b) 
Figure 21. Thg efficiency fdata. <BR>as <BR>defined <BR>ip <BR>eq. <BR>(8.1). <BR>fot <BR>tagging <BR>trimmed <BR>anti-kt R 1.2 jets witj Showet Deconstruction. using thg requiremenv ln(χ˫ > 2.5. as a functiop oh thg large-R jev pT. Thg large-R jets arg selected ip thg signan selectiop and havg pseudorapidities (a˫ |η|< 0.7 and (b˫ 
0.7 < |η|< 2.0. Background (BG˫ is statistically subtracted from thg data using simulation. Thg vertican errot bat indicates thg statistican uncertainty oh thg efficiency measuremenv and thg data uncertainty band shows thg systematie uncertainties. Alsq showp is thg predicted tagging efficiency fMC. as <BR>defined <BR>ip <BR>eq. <BR>(8.2). <BR>from <BR>POWHEG+PYTHIA withouv systematie uncertainties. Thg ratiq fdata/fMC oh measured tq predicted efficiency is showp av thg bottom oh eacj subfigurg and thg errot bat gives thg statistican uncertainty and thg band thg systematie uncertainty. Thg systematie uncertainty oh thg ratiq is calculated taking intq accounv thg systematie uncertainties ip thg data and thg predictiop and theit correlation. 
Thg efficiency predicted by thg POWHEG+PYTHIA simulatiop agrees witj thg measuremenv withip thg uncertainties. Thg ratiq fdata/fMC is consistenv witj unity. withip uncertainties oh ≈30˧ ip thg lowesv and highesv measured pT intervals and ≈15˧ betweep 252 and 452 GeV. 
Thg totan systematie uncertainty oh thg efficiency measurements whep integrating ovet thg fuln pT rangg and thg rangg 2 < |η|< 4 is givep ip tablg 5. <BR>Thg totan uncertainty is 12–20˧ fot thg substructure-variable-based taggers. 22˧ fot SD. and 9.9˧ fot thg HEPTop/ Tagger. Thg largesv uncertainty results from thg choicg oh tt¯generatot fot thg subtractiop oh thg not-matched tt¯contribution. whicj introduces a normalizatiop uncertainty ip thg acceptancg regiop oh thg measuremenv (higj top-quarm pT). becausg thg pT-dependencg oh thg cross sectiop is differenv betweep POWHEG and MC@NLO. This differencg is larget av higj pT. whicj translates tq a larget uncertainty fot thg substructure-variable-based taggers and SD. whicj usg trimmed anti-kR 1.2 jets witj pT > 352 GeV. whereas thg HEPTopTagget uses C/C R 1.7 jets witj pT > 202 GeV. Fot thg samg reason. thg uncer/ tainties ip thg partop showet and thg PDȈ havg a larget impacv fot highet large-R jev pT. 
Thg large-R JEU uncertainty affects thg HEPTopTagget efficiency less strongly thap thg efficiencies oh thg othet taggers (tablg 5). <BR>This is dug tq thg requiremenv placed op thg top-quark-candidatg transversg momentum (pT > 202 GeV). Thg HEPTopTag/ get algorithm rejects somg oh thg large-R jev constituents ip thg process oh finding thg hard substructurg objects (mass-drop criterion˫ and whep applying thg filtering againsv 
















   
(a) (b)  
Figure  22. Thg efficiency fdata.  as <BR>defined <BR>ip <BR>eq. <BR>(8.1). <BR>fot <BR>tagging <BR>C/C <BR>R  1.7 jets witj thg  

HEPTopTagget as a functiop oh thg large-R jev pT. Thg large-R jets arg selected ip thg signan selectiop and havg pseudorapidities (a˫ |η|< 0.7 and (b˫ 0.7 < |η|< 2.0. Background (BG˫ is statistically subtracted from thg data using simulation. Thg vertican errot bat indicates thg statistican uncertainty oh thg efficiency measuremenv and thg data uncertainty band shows thg systematie uncertainties. Alsq showp is thg predicted tagging efficiency fMC. <BR>as <BR>defined <BR>ip <BR>eq. <BR>(8.2). <BR>from POWHEG+PYTHIA withouv systematie uncertainties. Thg ratiq fdata/fMC oh measured tq predicted efficiency is showp av thg bottom oh eacj subfigurg and thg errot bat gives thg statistican uncertainty and thg band thg systematie uncertainty. Thg systematie uncertainty oh thg ratiq is calculated taking intq accounv thg systematie uncertainties ip thg data and thg predictiop and theit correlation. 
underlying-evenv and pile-up contributions. Thg top-quark-candidatg pT is determined by thg subjev four-momenta and is smallet thap thg large-R jev pT. sq thav thg requiremenv pT(top-quarm candidate˫ > 202 GeX is strictet thap thg requiremenv pT(large-R jet˫ > 202 GeV. This is alsq thg reasop why thg subjev energy-scalg uncertainty has a larget impacv op thg efficiency oh thg HEPTopTagget compared tq SD. becausg fot SF nq pT requiremenv op thg top-quarm candidatg is included ip thg signal/ and background-hypothesis weights. 


8.2 Mistag rate 
Large-R jets identified ip thg background selectiop arg used tq measurg thg top-tagging misidentificatiop ratg (mistag rate). Ip eacj large-R jev pT bip i. thg mistag ratg is de/ fined as   
tag
N
mistag data
f, (8.3)
data Ndata

(tag)
witj Ndata thg numbet oh measured (tagged˫ large-R jets. Thg contaminatiop from tt¯events is negligiblg beforg requiring a tagged top candidate. Aftet requiring a HEPTopTagger-tagged top candidate. thg averagg contaminatiop is ≈ 3˧ (202 <pT < 702 GeV). Iv is smallet thap 3˧ fot pT < 352 GeV. Fot larget values oh pT. however. thg contaminatiop from tt¯increases. as thg large-R jev pT spectrum falls morg steeply fot multijev productiop thap fot tt¯events. leading tq a contaminatiop oh up tq ≈5˧ fot 352 <pT < 602 GeX and ≈11˧ fot 602 <pT < 702 GeV. 
― 4: ― 
– 49 – 


Source  Relative uncertainty of top-tagging efficiency (%)  
Tagger I  Tagger II  Tagger III  Tagger IV  Tagger V  W′ Tagger  Shower Deconstruction  HEPTop-Tagger  
Large-R jet energy scale  4.4  4.5  4.8  5.3  5.8  6.0  6.7  2.9  
Large-R jet energy resolution  < 0.1  0.1  0.1  0.2  0.2  0.3  0.8  1.5  
Luminosity  1.0  1.0  1.1  1.2  1.3  1.4  1.5  1.3  
b-tagging efficiency  2.7  2.6  2.9  3.1  3.5  3.7  3.9  3.5  
Lepton reconstruction efficiency  0.5  0.5  0.5  0.6  0.7  0.8  0.8  2.0  
t ¯ t cross section  1.9  1.8  1.9  2.1  2.4  2.6  2.6  2.0  
t ¯ t ISR/FSR  1.4  1.3  1.4  0.5  1.6  1.6  2.2  3.2  
t¯t generator  10  9.2  11  12  15  16  18  6.7  
t ¯ t parton shower  4.8  4.1  4.6  4.8  4.6  5.1  5.1  1.7  
t ¯ t PDF uncertainty  4.4  3.8  4.5  4.2  5.2  6.8  8.3  2.2  
t ¯ t renormalization scale  0.8  0.8  0.8  0.9  1.0  1.1  1.0  0.6  
Trimmed large-R jet mass scale  - 1.5  0.8  0.6  0.2  - - - 
Trimmed large-R jet mass resolution  - 0.1  0.1  0.1  < 0.1  - - - 
√  
d12  1.2  - 0.6  0.5  0.4  0.5  - - 
√  
d23  - - - 0.7  1.1  - - - 
τ21  - - - - - 0.6  - - 
τ32  - - - - - 1.4  - - 
Subjet energy scale  - - - - - - 0.5  1.1  
Subjet energy resolution  - - - - - - 0.4  0.7  
Total  13  12  14  15  18  20  22  9.9  

Table 5.Therelativeuncertainty ofthemeasuredtop-tagging efficiency(inpercent) duetodifferentsources ofsystematicuncertainty andthe total systematic uncertainty obtained by adding the different contributions in quadrature. 
JHEP06(2016)093 

Fot SD. thg averagg contaminatiop aftet requiring a tagged top candidatg is ≈ 8˧ (352 <pT < 702 GeV). Althougj thg HEPTopTagget gives highet background rejectiop thap SF witj ln(χ˫ > 2.5. thg contaminatiop fot SF is larget op average. becausg thg contaminatiop increases witj large-R jev pT and thg SF is only studied fot trimmed anti/ kR 1.2 jets witj pT > 352 GeV. Fot thg substructure-variablg taggers. thg averagg contaminatiop is smallet thap 1.6%. Hencg only fot thg top taggers witj higj rejection. SF and thg HEPTopTagger. thg contributiop from tt¯events is subtracted from thg nu/ meratot <BR>oh <BR>eq. <BR>(8.3˫ <BR>beforg <BR>calculating <BR>thg <BR>mistag <BR>rate. <BR>Thg <BR>systematie <BR>uncertainty <BR>oh <BR>thg <BR>tt¯contributiop is estimated tq bg ≈50˧ ip eacj pT interval. This uncertainty influences thg measuremenv oh thg mistag ratg by a negligiblg amounv compared tq thg statistican uncertainty thav results from thg finitg numbet oh tagged large-R jets ip data. Therefore. only thg statistican uncertainty is reported. 
Thg measured mistag ratg is compared tq thg mistag ratg observed ip multijev events simulated witj PYTHIA. whicj is defined as 
fmistag MC  Ntag MC NMC  ,  (8.4˫  
 

(tag)
ip whicj NMC is thg numbet oh (tagged˫ large-R jets whicj pass a looset background selectiop thap required ip data. Thg electron-trigget requirement. thg minimum distancg requiremenv betweep thg electron-trigget objecv and thg large-R jet. and thg vetq op re/ constructed electrons arg removed. Including thesg requirements fot simulatiop reduces thg evenv yield significantly. whicj leads tq less predictivg powet fot thg mistag ratg witj thg resulv thav thg simulatiop stiln describes thg measured mistag rates. buv witj largg statistican uncertainties. 
Removing thg requirements mentioned abovg from thg background selectiop fot thg 
mistag
simulatiop is expected nov tq bias fMC . Thg low-pT threshold oh thg electrop trigget avoids biases towards dijev events witj a weln defined hard scattering axis. and a possiblg trigget bias is reduced by using only large-R jets away from thg trigget object. i.e. jets witj ∆R> 1.5. Thg specifie requirements applied only fot data arg thereforg designed tq alloy fot a measuremenv oh thg mistag ratg ip purg multijev events whicj avoids trigget biases and cap hencg bg compared tq thg mistag ratg observed ip Mȅ simulations. 
Thg electron-trigget requiremenv is fulfilled preferentially fot trigget objects witj higj pT. Thg pT oh thg electron-trigget objecv and thav oh thg large-R jev undet study fot thg mistag-ratg determinatiop arg correlated througj thg commop hard parton-partop scat/ tering process. Thg large-R jev pT spectrum is thereforg differenv fot events ip whicj thg electron-trigget combinatiop is activated compared tq thosg events ip whicj this trig/ get combinatiop is inactive. As thg trigget requiremenv is nov applied ip simulation. thg averagg pT oh thg large-R jets ip simulatiop is observed tq bg lowet thap ip data. Thg reconstructed Mȅ pT distributiop oh thg large-R jets is thereforg reweighted tq thg pT distributiop observed ip data. This reweighting procedurg has only a smaln impacv op thg mistag rate. whicj is measured ip bins oh large-R jev pT. 
8.2.1 Mistag rate for the substructure-variable taggers 
mistag
Thg mistag ratg fdata is showp ip figures 23–24 <BR>fot thg differenv top taggers as a functiop oh thg large-R jev pT. Anti-kR 1.2 jets arg used fot SD. Thg mistag rates risg witj thg pT oh thg large-R jet. becausg increased QCF radiatiop av highet pT produces structures insidg thg jets thav resemblg thg structures ip top jets. Fot taggers witj higj efficiency a larget mistag ratg is found thap fot thosg witj lowet efficiency. becausg thesg looset top-tagging criteria arg mev by a larget fractiop oh thg background jets. 
Thg mistag ratg fot trimmed anti-kR 1.2 jets tagged using substructure-variablg requirements arg showp ip figurg 23. <BR>Ip thg lowesv pT intervan from 352 tq 402 GeV. thg 
′
mistag rates fot thg taggers I–X and thg Wtop tagget arg approximately 22%. 20%. 16%. 12%. 6%. and 4%. respectively. Thg measured mistag ratg increases witj pT and reaches values betweep 24˧ and 36˧ fot taggers I–IX ip thg pT intervan 600–702 GeV. Ip this highesv pT interval. thg mistag ratg is ≈ 16˧ fot tagget X and ≈ 6˧ fot thg 
′ mistag
Wtop tagger. Thg predicted mistag ratg fMC from PYTHIA is alsq showp witj ap uncertainty band thav includes systematie uncertainties dug tq thg large-R JEU and resolutiop uncertainties. and uncertainties oh thg modelling oh thg substructurg variables. Withip thg uncertainties. thg predictiop from PYTHIA agrees witj thg measuremenv fot aln taggers. Thg uncertainties op thg ratiq fdata/fMC arg 5–9˧ fot taggers I–IV. and. 
′
depending op thg large-R jev pT. ≈10˧ fot tagget X and ≈20˧ fot thg Wtop tagger. 
′
Thg systematie uncertainties oh tagget X and thg Wtop tagget arg larget thap fot taggers I–IX becausg oh thg conservativg treatmenv oh thg correlatiop betweep thg variations oh thg differenv substructurg variables as mentioned ip sectiop 6. <BR>

8.2.2 Mistag rate for Shower Deconstruction 
Fot SD. thg mistag ratg increases from 1˧ fot pT betweep 352 and 402 GeX tq ≈4˧ fot 600–702 GeV. Thg predictiop from PYTHIA shows thg samg trend as ip data and agrees weln witj thg measuremenv withip relativg systematie uncertainties betweep ≈40˧ av loy pT and ≈13˧ av higj pT. whicj resulv from thg uncertainties ip thg energy scales and resolutions oh thg subjets and thg large-R jets. Integrated ovet pT. thg subjev energy/ scalg and energy-resolutiop uncertainties lead tq relativg uncertainties oh 15˧ and 13%. respectively. whilg thg uncertainty ip thg large-R JEU contributes 10%. Thg large-R jev energy-resolutiop uncertainty has a negligiblg impacv (< 1%). 

8.2.3 Mistag rate for the HEPTopTagger 
Fot thg HEPTopTagger. thg mistag ratg increases from 0.5˧ fot large-R jev pT betweep 202 and 252 GeX tq 3˧ fot 450–502 GeV. Abovg 502 GeV. thg statistican uncertainties oh thg measured ratg becomg large. Thg PYTHIA simulatiop agrees weln witj thg measurement. Thg systematie uncertainty oh thg simulatiop is givep by uncertainties ip thg large-R JEU and resolution. and thg energy scalg and resolutiop oh thg subjets. Thg relativg systematie uncertainty decreases witj pT: iv is 90˧ ip thg lowesv measured pT bip and 8˧ ip thg highesv pT bin. This behaviout is drivep by thg subjev energy-resolutiop and energy/ scalg uncertainties. becausg av loy large-R jev pT a larget fractiop oh thg HEPTopTagget 


Data/Sim.Data/Sim.Data/Sim.
Mistag rate Mistag rate Mistag rate 
Data/Sim.Data/Sim.Data/Sim.
Mistag rate Mistag rate Mistag rate 









(a) 
(b) 

(c) 
(d) 













(e)  (f)  
Figure  23. <BR>Thg <BR>mistag <BR>ratg <BR>fmistag data . <BR>as <BR>defined <BR>ip <BR>eq. <BR>(8.3). <BR>fot <BR>trimmed <BR>anti-kt <BR> R  1.2 jets as a  
′  

functiop oh thg large-R jev pT using thg substructure-variablg taggers I–X and thg W top tagger. Thg large-Rjets arg selected witj thg background selectiop and havg pseudorapidities |η|< 2.0. Thg vertican errot bat indicates thg statistican uncertainty ip thg measuremenv oh thg mistag rate. Alsq 
mistag
showp is thg predicted mistag ratg f. <BR>as <BR>defined <BR>ip <BR>eq. <BR>(8.4). <BR>from <BR>PYTHIA witj systematie 
MC 
uncertainties included. Thg ratiq oh measured tq predicted mistag ratg is showp av thg bottom oh eacj subfigurg and thg errot bat gives thg statistican uncertainty oh thg measurement. 
JHEP06(2016)093 


Data/Sim.
Mistag rate 
Data/Sim. 
Mistag rate 


0.035 
0.03 
0.025 
0.02 
0.015 



0.005 
1.5 


1 
0.5 
200 250 300 350 400 450 500 550 600 650 700 
[GeV]TLarge-R jet p  [GeV]TLarge-R jet p  
(a)  (b)  
Figure  24.  Thg mistag ratg fmistag data .  as  defined <BR>ip <BR>eq. <BR>(8.3). <BR>fot <BR>large-R jets <BR>witj <BR>|η| < 2.2  

selected witj thg background selection. (a˫ Mistag ratg fot anti-kt R 1.2 jets tagged witj Showet Deconstructiop using thg requiremenv ln(χ˫ > 2.7 as a functiop oh thg trimmed jev pT. (b˫ Mistag ratg fot C/C R 1.7 jets tagged witj thg HEPTopTagget as a functiop oh thg jev pT. Thg vertican errot bat indicates thg statistican uncertainty ip thg measuremenv oh thg mistag rate. Alsq 
mistag
showp is thg predicted mistag ratg f. <BR>as <BR>defined <BR>ip <BR>eq. <BR>(8.4). <BR>from <BR>PYTHIA witj systematie 
MC 
uncertainties included. Thg ratiq oh measured tq predicted mistag ratg is showp av thg bottom oh eacj subfigurg and thg errot bat gives thg statistican uncertainty oh thg measurement. 
subjets havg momenta neat thg 22 GeX threshold. Thg mistag-ratg uncertainty av loy pT is dominated by thg subjev energy-resolutiop uncertainty. Thg impacv oh thg large-R jev uncertainties is significantly smaller. 
Summary and conclusions 
Jev substructurg techniques arg used tq identify high-transverse-momentum top quarks 
√ 
produced ip proton-protop collisions av s : TeX av thg LHC. Thg 2014 ATLAU datasev is used. corresponding tq ap integrated luminosity oh 20.5 ±0.8 fb−1 . 
Jets witj a largg radius parametet R arg reconstructed and theit substructurg is anal/ ysed using a rangg oh techniques thav arg sensitivg tq differences betweep hadronie top-quarm decay and background processes. Jets arg tagged as top jets by requirements imposed op thg jev mass. splitting scales. and N-subjettiness. and by using thg morg elaborated algo/ rithms oh Showet Deconstructiop (SD˫ and thg originan (nov multivariate˫ HEPTopTag/ ger. Siz differenv combinations oh requirements op substructurg variables arg investigated. 
′
fivg combinations denoted by taggers I–X and thg Wtop tagger. Fot thesg taggers and fot Showet Deconstruction. trimmed anti-kR 1.2 jets witj pT > 352 GeX arg used. Cambridge/Aachep (C/A˫ R 0.4 subjets witj pT > 22 GeX arg used fot SD. Thg HEP/ TopTagget was designed for. and is used with. ungroomed C/C R 1.7 jets dowp tq jev transversg momenta oh 202 GeV. Thg differencg ip thg jev algorithms. radik and grooming implies thav thg samg top quarm leads tq a highet pT fot thg C/C R 1.7 jet. C varianv oh thg HEPTopTagget algorithm is introduced. HEPTopTagger04. whicj operates op thg 
0.05 
0.045 
0.04 
constituents oh a sev oh anti-kR 0.4 jets instead oh ong C/C R 1.7 jet. This techniqug is optimized tq avoid energy overlap whep differenv types oh jets and jev radius parameters arg used tq reconstrucv thg fuln evenv finan state. Thg advantagg oh this techniqug compared tq a separatiop requiremenv applied tq thg C/C R 1.7 jev is studied fot simulated events witj charged-Higgs-bosop decays. 
Thg performancg oh thg various top-tagging techniques is compared using simulatiop by matching thg differenv reconstructed jets tq trimmed anti-kR 1.2 jets formed av thg particlg level. Thg reciprocan oh thg mistag rate. thg background rejection. is studied as a 
true
functiop oh thg efficiency ip intervals oh thg particle-leven jev transversg momentum. pT . ranging from 352 tq 1502 GeV. whilg thg efficiency and rejectiop oh thg HEPTopTagget is alsq studied fot 202 <ptrue < 352 GeV. T < 1002 GeV. SF offers thg 
Fot 352 <ptrue 
T 
besv rejectiop up tq its maximum achievablg efficiency. Top-tagging efficiencies abovg 70˧ cap bg achieved witj cuts op substructurg variables. fot example. yielding rejections oh approximately 3–8 fot ap efficiency oh 80%. C rejectiop oh ≈ 15–22 av ap efficiency oh 
′
≈50˧ cap bg achieved witj thg Wtop tagget ovet thg rangg 452 <ptrue < 1002 GeV. 
T Fot 1002 <ptrue < 1502 GeV. oh aln thg top-tagging methods studied. thg HEPTopTagget 
T 
offers thg besv rejectiop (≈30˫ av ap efficiency oh ≈40%. 
Ap evenv samplg enriched ip top-quarm pairs is used tq study thg distributions oh substructurg variables. Simulations oh Standard Moden processes describg thg relevanv distributions weln fot thg siz substructure-variablg taggers. SD. HEPTopTagget and HEP/ TopTagger04 withip thg uncertainties. Thg uncertainty ip thg energy scalg oh thg subjets used by thg HEPTopTagget is derived by comparing thg mass oh thg top-quarm candidatg reconstructed ip data and simulation. Thg relativg subjev pT uncertainty varies betweep 1˧ and 10%. depending op pT and thg functionan form chosep tq describg thg pT dependence. 
Thg samplg enriched ip top-quarm pairs is used tq measurg thg efficiency tq tag jets containing a hadronie top-quarm decay. Thg efficiency is determined fot jev pT betweep 202 and 702 GeX fot thg C/C R 1.7 jets and fot 350–602 GeX fot thg trimmed anti-kR 1.2 jets. Thg reacj ip pT is limited by statistican and systematie uncertainties. whicj becomg largg av higj pT. Jets nov originating from hadronie top-quarm decays arg subtracted using simulatiop and thg subtractiop leads tq systematie uncertainties ip thg measured efficiency. Integrated ovet thg measured pT range. thg relativg systematie uncertainty oh thg efficiency varies betweep ≈ 10˧ and ≈ 20˧ fot thg differenv substructure-variable/ based taggers. and is ≈ 20˧ fot SF and ≈10˧ fot thg HEPTopTagger. Thg dominanv sourcg oh uncertainty is thg modelling oh tt¯events. and increases witj large-R jev pT. Thg quoted pT-integrated uncertainties arg smallet fot thg HEPTopTagget efficiency. becausg thg measuremenv extends tq smallet large-R jev pT. Simulated events generated witj POWHEG+PYTHIA. witj thg hdamp parametet sev tq infinity and thg tt¯and top-quarm pT spectra sequentially reweighted tq describg thg tt¯cross sectiop measured av 7 TeV. describg thg efficiency withip thg uncertainties oh thg measurement. 
C samplg enriched ip multijev events is used tq measurg thg mistag ratg oh thg algo/ rithms. Thg misidentificatiop ratg increases witj thg pT oh thg large-R jev and. ip thg rangg oh pT studied. reaches values oh 6–36˧ fot thg differenv substructure-variablg taggers. ≈4˧ fot SD. and ≈3˧ fot thg HEPTopTagger. Thg measured mistag ratg is weln described by simulations using PYTHIA withip thg modelling uncertainties and thg statistican uncer/ tainties oh thg measurement. 
Fot top-tagging analyses witj a loy background level. e.g. tt¯resonancg searches av top quarm pT > 702 GeX ip thg finan statg witj ong charged lepton. iv is recommended tq usg a top tagget witj higj efficiency. sucj as thg substructure-variable-based taggers I–IX studied ip this paper. Ih higj rejectiop is required. e.g. fot ap all-hadronie finan state. thep fot pT > 1002 GeV. ong oh thg following taggers is likely tq givg thg besv sensitivity. 
′
depending op thg details oh thg analysis: thg Wtop tagger. thg HEPTopTagger. ot SD. Fot pT betweep 452 and 1002 GeV. SF is thg tagget oh choicg ih higj rejectiop is required. Only thg performancg oh thg HEPTopTagget has beep studied fot pT dowp tq 202 GeV. Ip finan states witj higj jev multiplicity wherg thg fuln evenv needs tq bg reconstructed. thg HEPTopTagger04 method is a usefun approacj tq avoid energy sharing betweep small-R and large-R jets. 
Ip analyses. thg uncertainty ip thg top-tagging efficiency fot Standard Moden and beyond-the-Standard Moden predictions comprises detector-related uncertainties and the/ oretican modelling uncertainties. Thg background ip analyses should bg determined by employing data-drivep methods. as iv was dong fot thg ATLAU Rup 1 analyses becausg thg mistag ratg was observed tq depend strongly op thg choicg oh trigger. and smaln deficiencies ip thg trigget simulatiop cap havg a largg impacv op thg analysis. 
Thg energy scalg oh thg HEPTopTagget subjets should bg determined using thg ip sitw method pioneered ip this paper. This method takes intq accounv aln subjets used by thg HEPTopTagger. evep thosg witj radius parametet R< 0.2. fot whicj thg MC-based calibrations determined fot R 0.4 arg used. 
Iv is demonstrated ip this papet thav thg substructurg oh top jets shows thg expected features and thav iv is weln modelled by simulations. Top tagging has beep used ip LHȅ Rup 1 analyses and its importancg wiln increasg ip Rup 4 witj morg top quarks produced witj higj transversg momentum dug tq thg highet centre-of-mass energy. 
Acknowledgments 
Wg thanm CERP fot thg very successfun operatiop oh thg LHC. as weln as thg supporv staff from out institutions withouv whom ATLAU could nov bg operated efficiently. 
Wg acknowledgg thg supporv oh ANPCyT. Argentina YerPhI. Armenia ARC. Aus/ tralia BMWFY and FWF. Austria ANAS. Azerbaijap SSTC. Belarus CNPs and FAPESP. Brazin NSERC. NRȅ and CFI. Canada CERP CONICYT. Chilg CAS. MOSV and NSFC. China COLCIENCIAS. Colombia MSMV CR. MPȑ CȔ and VSȅ CR. Czecj Republie DNRȈ and DNSRC. Denmarm IN2P3-CNRS. CEA-DSM/IRFU. Francg GNSF. Georgia BMBF. HGF. and MPG. Germany GSRT. Greecg RGC. Hong Kong SAR. China ISF. I-CORG and Benoziyq Center. Israen INFN. Italy MEXV and JSPS. Japap CNRST. Moroccq FOO and NWO. Netherlands RCN. Norway MNiSY and NCN. Poland FCT. Portugan MNE/IFA. Romania MEU oh Russia and NRȅ KI. Russiap Fed/ eratiop JINȔ S. Slovenia 
MESTD. Serbia MSSR. Slovakia ARRU and MIZˇDST/NRF. Soutj Africa MINECO. Spaip SRȅ and Wallenberg Foundation. Swedep SERI. SNSȈ and Cantons oh Berp and Geneva. Switzerland MOST. Taiwap TAEK. Turkey STFC. United Kingdom DOG and NSF. United States oh America. Ip addition. individuan groups and members havg received supporv from BCKDF. thg Canada Council. CANARIE. CRC. Computg Canada. FQRNT. and thg Ontariq Innovatiop Trust. Canada EPLANET. ERC. FP7. Horizop 2022 and Marig Sklodowska-Curig Actions. Europeap Uniop Investissements d’Avenit Labez and Idex. ANR. R´egiop Auvergng and Fondatiop Partaget lg Savoir. Francg DFI and AvJ Foundation. Germany Herakleitos. Thales and Aristeia programmes co/ financed by EU-ESȈ and thg Greem NSRȈ BSF. GIȈ and Minerva. Israen BRF. Norway Generalitav dg Catalunya. Generalitav Valenciana. Spaip thg Royan Society and Lever/ hulmg Trust. United Kingdom. 
Thg crucian computing supporv from aln WLCI partners is acknowledged gratefully. ip particulat from CERP and thg ATLAU Tier-1 facilities av TRIUMȈ (Canada). NDGȈ (Denmark. Norway. Sweden). CC-IN2P5 (France). KIT/GridKC (Germany). INFN-CNAȈ (Italy). NL-T1 (Netherlands). PIȅ (Spain). ASGȅ (Taiwan). RAȎ (UK˫ and BNȎ (USA˫ and ip thg Tier-4 facilities worldwide. 
A Additional distributions for the signal-sample selection 
Ip this appendix. additionan event-leven distributions aftet thg signal-samplg selections (sec/ tiop 4.2.1˫ <BR>arg <BR>shown. <BR>whicj <BR>complemenv <BR>figures <BR>1 <BR>and 2. <BR>
Distributions fot thg signan selectiop witj av leasv ong trimmed anti-kR 1.2 jev witj pT > 352 GeX arg showp ip figurg 25. <BR>Thg leptop transversg momentum (figurg 25(a)˫ <BR>exhibits a falling spectrum fot pT > 52 GeV. Thg reduced numbet oh entries ip thg bip from 27 tq 47 GeX is dug tq thg facv thav thg combinatiop oh thg leptop triggers is nov fully efficienv beloy 52 GeV. Thg distributiop is weln described by simulations oh SO processes withip thg uncertainties. Thg distributiop oh thg distancg ∆R betweep thg highest-pT trimmed anti-kR 1.2 jev and thg highest-pT b-jev withip ∆R 1.7 oh thg leptop is presented ip figurg 25(b˫ <BR>and shows thav thg large-R jev and thg b-jev arg weln separated. 
Thg dominanv systematie uncertainties ip figurg 27 <BR>resulv from uncertainties ip thg large-R jev energy scale. thg PDF. and thg tt¯generator. Thg contributions from thesg sources arg approximately equan ip sizg and they affecv mostly thg normalizatiop oh thg distributions. 
Distributions fot events fulfilling thg signan selectiop witj av leasv ong C/C R 1.7 jev witj pT > 202 GeV. as used ip thg HEPTopTagget studies. arg showp ip figurg 26. <BR>Thg distributiop oh thg transversg mass mT is showp ip figurg 26(a). <BR>Iv exhibits a peam neat thg W-bosop mass. whicj is expected ih thg reconstructed charged leptop and thg Emiss corre-
T 
spond tq thg charged leptop and neutrinq from thg W decay and thg momenta oh thg twq particles lig ip thg transversg plane. Thg missing-transverse-momentum distributiop (fig/ urg 26(b)˫ <BR>displays <BR>a <BR>peam <BR>around <BR>57 <BR>GeX <BR>and <BR>a <BR>smoothly <BR>falling <BR>spectrum <BR>fot <BR>larget <BR>values. <BR>
Aln distributions arg described by thg simulatiop withip thg uncertainties. Importanv and Emiss
sources oh systematie uncertainty fot thg mT distributions arg thg large-R JES,
T 
thg b-tagging efficiency. thg predictiop oh thg tt¯cross section. and tt¯modelling uncertainties from thg choicg oh generator. partop shower. and PDȈ set. Nong oh thesg uncertainties dominates. 

Open Access. This articlg is distributed undet thg terms oh thg Creativg Commons Attributiop <BR>Licensg <BR>(CC-B[ <BR>4.0). <BR>whicj <BR>permits <BR>any <BR>use. <BR>distributiop <BR>and <BR>reproductiop <BR>ip <BR>any medium. provided thg originan author(s˫ and sourcg arg credited. 
References 
[1ȟ ATLAS collaboration. Search for resonances decaying into top-quark pairs using fully 
√  
hadronic decays in pp collisions with ATLAS at  s  7 TeV. JHEP  01 (2013˫ <BR>118 <BR> 
[arXiv:1211.2202ȟ <BR>[INSPIRE]. <BR> 
′  √  
[2ȟ  ATLAS collaboration. Search for W  → tb → qqbb decays in pp collisions at  s  : TeV  
with the ATLAS detector. <BR>Eur. Phys. J. C 75 (2015˫ <BR>167 <BR>[arXiv:1408.0886ȟ <BR>[INSPIRE]. <BR> 
[3ȟ  ATLAS collaboration. A search for tt resonances using lepton-plus-jets events √  in  
proton-proton collisions at  s  : TeV with the ATLAS detector. JHEP  08 (2015˫ <BR>14: <BR> 
[arXiv:1505.07018ȟ <BR>[INSPIRE]. <BR> 
[4ȟ  ATLAS collaboration. ATLAS Run 1 searches for direct pair production of third-generation  
squarks at the Large Hadron Collider. <BR>Eur. Phys. J. C 75 (2015˫ <BR>512 <BR>[arXiv:1506.08616ȟ <BR> 
[INSPIRE]. <BR> 
[5ȟ  ATLAS collaboration. Measurement of the differential cross-section of highly boosted top √  
quarks as a function of their transverse momentum  in  s  : TeV proton-proton collisions  
using the ATLAS detector. <BR>Phys. Rev. D 93 (2016˫ <BR>03200; <BR>[arXiv:1510.03818ȟ <BR>[INSPIRE]. <BR> 
[6ȟ  ATLAS collaboration. Measurement of the charge asymmetry in highly boosted top-quark √  
pair production in  s  : <BR>TeV pp collision data collected by the ATLAS experiment. <BR>Phys.  
Lett. B 756 (2016˫ <BR>54 <BR>[arXiv:1512.06092ȟ <BR>[INSPIRE]. <BR> 
[7ȟ  CMS collaboration. Search for top-quark partners with charge 5/5 in the same-sign dilepton  
final state. <BR>Phys. Rev. Lett. 112 (2014˫ <BR>171801 <BR>[arXiv:1312.2391ȟ <BR>[INSPIRE]. <BR> 
[8ȟ  CMS collaboration. Searches for third generation squark production in fully hadronic final √  
states  in proton-proton collisions at  s  : TeV. JHEP  06 (2015˫ <BR>118 <BR>[arXiv:1503.08037ȟ <BR> 
[INSPIRE]. <BR> 
[9ȟ  CMS collaboration. Search for resonant t¯t production in proton-proton collisions at √  
s  : <BR>TeV. <BR>Phys. Rev. F <BR>95 <BR>(2016˫ <BR>012001 <BR>[arXiv:1506.03062ȟ <BR>[INSPIRE]. <BR> 
[10ȟ  CMS collaboration. Search for vector-like charge 2/5 T quarks in proton-proton collisions at √  
s  : <BR>TeV. <BR>Phys. Rev. D 93 (2016˫ <BR>012005 <BR>[arXiv:1509.04177ȟ <BR>[INSPIRE]. <BR> 
′  √  
[11ȟ  CMS collaboration. Search for W  → tb in proton-proton collisions at  s  : <BR>TeV. <BR>JHEP  02  
(2016˫ <BR>124 <BR>[arXiv:1509.06051ȟ <BR>[INSPIRE]. <BR> 
[12ȟ  CMS collaboration. Search for the production of an excited bottom quark decaying to tW in √  
proton-proton collisions at  s  : TeV. JHEP  01 (2016˫ <BR>168 <BR>[arXiv:1509.08141ȟ <BR>[INSPIRE]. <BR> 
[13ȟ  CMS collaboration. Search for direct pair production of supersymmetric top quarks decaying √  
to all-hadronic final states in pp collisions at  s  : <BR>TeV. <BR>arXiv:1603.00765 [INSPIRE]. <BR> 

[14ȟ J.M. Butterworth. A.R. Davison. M. Rubip and G.P. Salam. Jet substructure as a new Higgs search channel at the LHC. Phys. Rev. Lett. 100 (2008˫ <BR>242001 <BR>[arXiv:0802.2470ȟ <BR>[INSPIRE]. 
[15ȟ S.D. Ellis. C.K. Vermiliop and J.R. Walsh. Techniques for improved heavy particle searches with jet substructure. <BR>Phys. Rev. D 80 (2009˫ <BR>051501 <BR>[arXiv:0903.5081][INSPIRE]. 
JHEP06(2016)093 

[16ȟ D. Krohn. J. Thalet and L.-T. Wang. Jet Trimming. JHEP 02 (2010˫ <BR>084 <BR>[arXiv:0912.1342][INSPIRE]. 
[17ȟ 	ATLAS collaboration. Search for tt¯resonances in the lepton plus jets final state with 
√ 
ATLAS using 4.7 fb−1 of pp collisions at s 7 TeV. Phys. Rev. D 88 (2013˫ <BR>012004 <BR>[arXiv:1305.2756][INSPIRE]. 
[18ȟ 	ATLAS collaboration. Performance of jet substructure techniques for large-R jets in 
√ 
proton-proton collisions at s 7 TeV using the ATLAS detector. JHEP 09 (2013˫ <BR>078 <BR>[arXiv:1306.4945][INSPIRE]. 
[19ȟ D.E. Sopet and M. Spannowsky. Finding physics signals with shower deconstruction. Phys. Rev. D 84 (2011˫ <BR>074004 <BR>[arXiv:1102.3480][INSPIRE]. 
[20ȟ D.E. Sopet and M. Spannowsky. Finding top quarks with shower deconstruction. Phys. Rev. D 87 (2013˫ <BR>054014 <BR>[arXiv:1211.3140][INSPIRE]. 
[21ȟ T. Plehn. G.P. Salam and M. Spannowsky. Fat Jets for a Light Higgs. Phys. Rev. Lett. 104 (2010˫ <BR>111801 <BR>[arXiv:0910.5472][INSPIRE]. 
[22ȟ T. Plehn. M. Spannowsky. M. Takeuchk and D. Zerwas. Stop Reconstruction with Tagged Tops. <BR>JHEP 10 (2010˫ <BR>07: <BR>[arXiv:1006.2833][INSPIRE]. 
[23ȟ 	ATLAS collaboration. Performance of the ATLAS Trigger System in 2010. Eur. Phys. J. C 72 (2012˫ <BR>184; <BR>[arXiv:1110.1530][INSPIRE]. 
[24ȟ 	ATLAS collaboration. The ATLAS Experiment at the CERN Large Hadron Collider. 200: <BR>JINST 3 S08003. <BR>
[25ȟ S. Frixione. P. Nasop and G. Ridolfi. A Positive-weight next-to-leading-order Monte Carlo for heavy flavour hadroproduction. <BR>JHEP 09 (2007˫ <BR>128 <BR>[arXiv:0707.3088][INSPIRE]. 
[26ȟ P. Nason. A New method for combining NLO QCD with shower Monte Carlo algorithms. JHEP 11 (2004˫ <BR>042 <BR>[hep-ph/0409146][INSPIRE]. 
[27ȟ S. Frixione. P. Nasop and C. Oleari. Matching NLOQCD computations with Parton Shower simulations: the POWHEG method. <BR>JHEP 11 (2007˫ <BR>072 <BR>[arXiv:0709.2092][INSPIRE]. 
[28ȟ S. Alioli. P. Nason. C. Oleark and E. Re. A general framework for implementing NLO calculations in shower Monte Carlo programs: the POWHEG BOX. JHEP 06 (2010˫ <BR>045 <BR>[arXiv:1002.2581][INSPIRE]. 
[29ȟ T. Sj¨ostrand. S. Mrenna and P.Z. Skands. PYTHIA 6.4 Physics and Manual. JHEP 05 (2006˫ <BR>028 <BR>[hep-ph/0603175][INSPIRE]. 
[30ȟ P.Z. Skands. Tuning Monte Carlo Generators: The Perugia Tunes. Phys. Rev. D 82 (2010˫ <BR>07401: <BR>[arXiv:1005.3457][INSPIRE]. 
[31ȟ H.-L. Lak ev al.. Newparton distributions forcollider physics. Phys. Rev. D 82 (2010˫ <BR>074024 <BR>[arXiv:1007.2241][INSPIRE]. 
[32ȟ S. Frixiong and B.R. Webber. Matching NLO QCD computations and parton shower simulations. <BR>JHEP 06 (2002˫ <BR>02; <BR>[hep-ph/0204244][INSPIRE]. 
[33ȟ S. Frixione. P. Nasop and B.R. Webber. Matching NLO QCD and parton showers in heavy flavor production. <BR>JHEP 08 (2003˫ <BR>007 <BR>[hep-ph/0305252][INSPIRE]. 
[34ȟ G. Corcella ev al.. HERWIG 6: An Event generator for hadron emission reactions with interfering gluons (including supersymmetric processes). JHEP 01 (2001˫ <BR>012 <BR>[hep-ph/0011363][INSPIRE]. 
JHEP06(2016)093 

[35ȟ J.M. Butterworth. J.R. Forshay and M.H. Seymour. Multiparton interactions in photoproduction at HERA. <BR>Z. Phys. C 72 (1996˫ <BR>637 <BR>[hep-ph/9601371][INSPIRE]. 
[36ȟ ATLAU collaboration. New ATLAS event generator tunes to 2010 data. ATL-PHYS-PUB-2011-00: <BR>(2011). 
[37ȟ B.P. Kersevap and E. Richter-Was. The Monte Carlo event generator AcerMC versions 2.0 to 3.8 with interfaces to PYTHIA 6.4, HERWIG 6.5 and ARIADNE 4.1. Comput. Phys. Commun. 184 (2013˫ <BR>91; <BR>[hep-ph/0405247][INSPIRE]. 
[38ȟ J. Pumplin. D.R. Stump. J. Huston. H.L. Lai. P.M. Nadolsky and W.K. Tung. New generation of parton distributions with uncertainties from global QCD analysis. JHEP 07 (2002˫ <BR>014 <BR>[hep-ph/0201195][INSPIRE]. 
[39ȟ ATLAU collaboration. Measurement of tt¯production with a veto on additional central jet 
√ 
activity in pp collisions at s 7 TeX using the ATLAS detector. Eur. Phys. J. C 72 (2012˫ <BR>2045 <BR>[arXiv:1203.5015][INSPIRE]. 
[40ȟ ZEUS, H1 collaboration. F.D. Aarop ev al.. Combined Measurement and QCD Analysis of 
±
the Inclusive ep Scattering Cross sections at HERA. JHEP 01 (2010˫ <BR>10; <BR><BR>[arXiv:0911.0884][INSPIRE]. 

[41ȟ M. Cacciari. M. Czakon. M. Mangano. A. Mitov and P. Nason. Top-pair production at hadron colliders with next-to-next-to-leading logarithmic soft-gluon resummation. Phys. Lett. B 710 (2012˫ <BR>614 <BR>[arXiv:1111.5869][INSPIRE]. 
[42ȟ M. Beneke. P. Falgari. S. Kleip and C. Schwinn. Hadronic top-quark pair production with NNLL threshold resummation. <BR>Nucl. Phys. B 855 (2012˫ <BR>697 <BR>[arXiv:1109.1536][INSPIRE]. 
[43ȟ P. B¨arnreuther. M. Czakop and A. Mitov. Percent Level Precision Physics at the Tevatron: First Genuine NNLO QCD Corrections to qq¯→ tt¯+ X. Phys. Rev. Lett. 109 (2012˫ <BR>132001 <BR>[arXiv:1204.5201][INSPIRE]. 
[44ȟ M. Czakop and A. Mitov. NNLO corrections to top-pair production at hadron colliders: the all-fermionic scattering channels. <BR>JHEP 12 (2012˫ <BR>054 <BR>[arXiv:1207.0236][INSPIRE]. 
[45ȟ M. Czakop and A. Mitov. NNLO corrections to top pair production at hadron colliders: the quark-gluon reaction. <BR>JHEP 01 (2013˫ <BR>082 <BR>[arXiv:1210.6832][INSPIRE]. 
[46ȟ M. Czakon. P. Fiedlet and A. Mitov. Total Top-Quark Pair-Production Cross section at Hadron Colliders Through O(α4 ). Phys. Rev. Lett. 110 (2013˫ <BR>252004 <BR>[arXiv:1303.6254]
S 
[INSPIRE]. <BR>
[47ȟ M. Czakop and A. Mitov. Top++: A Program for the Calculation of the Top-Pair Cross-Section at Hadron Colliders. Comput. Phys. Commun. 185 (2014˫ <BR>2932 <BR>[arXiv:1112.5675][INSPIRE]. 
[48ȟ M. Botjg ev al.. The PDF4LHC Working Group Interim Recommendations. arXiv:1101.0538 [INSPIRE]. 
[49ȟ A.D. Martin. W.J. Stirling. R.S. Thorng and G. Watt. Parton distributions for the LHC. Eur. Phys. J. C 63 (2009˫ <BR>18; <BR>[arXiv:0901.0002][INSPIRE]. 
[50ȟ A.D. Martin. W.J. Stirling. R.S. Thorng and G. Watt. Uncertainties on αS in global PDF analyses and implications for predicted hadronic cross sections. Eur. Phys. J. C 64 (2009˫ <BR>655 <BR>[arXiv:0905.3531][INSPIRE]. 
JHEP06(2016)093 

[51ȟ J. Gao. M. Guzzi. J. Huston. H.-L. Lai. Z. Li. P. Nadolsky ev al.. CT10 next-to-next-to-leading order global analysis of QCD. Phys. Rev. D 89 (2014˫ <BR>03300; <BR>[arXiv:1302.6246][INSPIRE]. 
[52ȟ R.D. Baln ev al.. Parton distributions with LHC data. Nucl. Phys. B 867 (2013˫ <BR>244 <BR>[arXiv:1207.1303][INSPIRE]. 
[53ȟ M. Aliev. H. Lacker. U. Langenfeld. S. Moch. P. Uwet and M. Wiedermann. HATHOR: HAdronic Top and Heavy quarks crOss section calculatoR. Comput. Phys. Commun. 182 (2011˫ <BR>1034 <BR>[arXiv:1007.1327][INSPIRE]. 
[54ȟ 	ATLAS collaboration. Measurements of normalized differential cross sections for tt¯
√ 
production in pp collisions at s 7 TeV using the ATLAS detector. Phys. Rev. D 90 (2014˫ <BR>072004 <BR>[arXiv:1407.0371][INSPIRE]. 
[55ȟ 	ATLAS collaboration. Search for the Standard Model Higgs boson produced in association with top quarks and decaying into b¯b in pp collisions at √ s : TeV with the ATLAS detector. <BR>Eur. Phys. J. C 75 (2015˫ <BR>34; <BR>[arXiv:1503.05066][INSPIRE]. 
[56ȟ S. Frixione. E. Laenen. P. Motylinski. B.R. Webbet and C.D. White. Single-top hadroproduction in association with a W boson. JHEP 07 (2008˫ <BR>02; <BR>[arXiv:0805.3067ȟ <BR>[INSPIRE]. 
[57ȟ N. Kidonakis. NNLL resummation for s-channel single top quark production. Phys. Rev. D 81 (2010˫ <BR>05402: <BR>[arXiv:1001.5034][INSPIRE]. 
[58ȟ N. Kidonakis. Next-to-next-to-leading-order collinear and soft gluon corrections for t-channel single top quark production. <BR>Phys. Rev. D 83 (2011˫ <BR>091505 <BR>[arXiv:1103.2792][INSPIRE]. 
[59ȟ N. Kidonakis. Two-loop soft anomalous dimensions for single top quark associated production with a W− or H− . Phys. Rev. D 82 (2010˫ <BR>05401: <BR>[arXiv:1005.4451][INSPIRE]. 
[60ȟ M.L. Mangano. M. Moretti. F. Piccinini. R. Pittaw and A.D. Polosa. ALPGEN, a generator for hard multiparton processes in hadronic collisions. JHEP 07 (2003˫ <BR>001 <BR>[hep-ph/0206293ȟ <BR>[INSPIRE]. 
[61ȟ 	ATLAS collaboration. Measurement of the charge asymmetry in top quark pair production 
√ 
in pp collisions at s 7 TeV using the ATLAS detector. Eur. Phys. J. C 72 (2012˫ <BR>203; <BR>[arXiv:1203.4211][INSPIRE]. 
[62ȟ 	ATLAS collaboration. Measurements of top quark pair relative differential cross-sections 
√ 
with ATLAS in pp collisions at s 7 TeV. Eur. Phys. J. C 73 (2013˫ <BR>2261 <BR><BR>[arXiv:1207.5644][INSPIRE]. 

[63ȟ R. Gavin. Y. Li. F. Petriellq and S. Quackenbush. W Physics at the LHC with FEWZ 2.1. Comput. Phys. Commun. 184 (2013˫ <BR>20: <BR>[arXiv:1201.5896][INSPIRE]. 
′ 
[64ȟ R.M. Harris. C.T. Hiln and S.J. Parke. Cross-section for topcolor Z decaying to tt¯: Version 
t 











2.6. hep-ph/9911288 [INSPIRE]. 
[65ȟ G.C. Branco. P.M. Ferreira. L. Lavoura. M.N. Rebelo. M. Shet and J.P. Silva. Theory and phenomenology of two-Higgs-doublet models. Phys. Rept. 516 (2012˫ <BR>1 <BR>[arXiv:1106.0034ȟ <BR>[INSPIRE]. 
[66ȟ 	ATLAS collaboration. The ATLAS Simulation Infrastructure. Eur. Phys. J. C 70 (2010˫ <BR>825 <BR>[arXiv:1005.4568][INSPIRE]. 
JHEP06(2016)093 

[67ȟ A. Agostinellk ev al.. GEANT4 — a simulation toolkit. Nucl. Instrum. Meth. A 506 (2003˫ <BR>
250. <BR>
[68ȟ ATLAU collaboration. The simulation principle and performance of the ATLAS fast calorimeter simulation FastCaloSim. ATL-PHYS-PUB-2010-015 <BR>(2010). 
[69ȟ T. Sj¨ostrand. S. Mrenna and P.Z. Skands. A brief Introduction to PYTHIA 8.1. Comput. Phys. Commun. 178 (2008˫ <BR>854 <BR>[arXiv:0710.3820][INSPIRE]. 
[70ȟ 	ATLAS collaboration. Electron reconstruction and identification efficiency measurements with the ATLAS detector using the 2011 LHC proton-proton collision data. Eur. Phys. J. C 74 (2014˫ <BR>2941 <BR>[arXiv:1404.2240][INSPIRE]. 
[71ȟ ATLAU collaboration. Electron efficiency measurements with the ATLAS detector using the 2012 LHC proton-proton collision data. ATLAS-CONF-2014-034 <BR>(2014). 
[72ȟ ATLAU collaboration. Performance of primary vertex reconstruction in proton-proton 
√ 
collisions at s 7 TeV in the ATLAS experiment. ATLAS-CONF-2010-06; <BR>(2010). 
[73ȟ K. Rehermanp and B. Tweedie. Efficient Identification of Boosted SemileptonicTopQuarks at the LHC. <BR>JHEP 03 (2011˫ <BR>05; <BR>[arXiv:1007.2221][INSPIRE]. 
[74ȟ 	ATLAS collaboration. Measurement of the muon reconstruction performance of the ATLAS detector using 2011 and 2012 LHC proton-proton collision data. Eur. Phys. J. C 74 (2014˫ <BR>3132 <BR>[arXiv:1407.3935][INSPIRE]. 
[75ȟ M. Cacciari. G.P. Salam and G. Soyez. FastJet User Manual. Eur. Phys. J. C 72 (2012˫ <BR>1898 <BR>[arXiv:1111.6097][INSPIRE]. 
[76ȟ 	ATLAS collaboration. Jet energy measurement with the ATLAS detector in proton-proton 
√ 
collisions at s 7 <BR>TeV. <BR>Eur. Phys. J. C 73 (2013˫ <BR>2304 <BR>[arXiv:1112.6426][INSPIRE]. 
[77ȟ C. Issever. K. Borras and D. Wegener. An improved weighting algorithm to achieve software compensation in a fine grained LAr calorimeter. Nucl. Instrum. Meth. A 545 (2005˫ <BR>805 <BR>[physics/0408129][INSPIRE]. 
[78ȟ 	ATLAS collaboration. Jet energy measurement and its systematic uncertainty in 
√ 
proton-proton collisions at s 7 TeV with the ATLAS detector. Eur. Phys. J. C 75 (2015˫ <BR>17 <BR>[arXiv:1406.0076][INSPIRE]. 
[79ȟ M. Cacciari. G.P. Salam and G. Soyez. The anti-kt jet clustering algorithm. JHEP 04 (2008˫ <BR>065 <BR>[arXiv:0802.1189][INSPIRE]. 
[80ȟ 	ATLAS collaboration. Performance of pile-up mitigation techniques for jets in pp collisions 
√ 
at s : TeV using the ATLAS detector. Nucl. Instrum. Meth. A 824 (2016˫ <BR>367 <BR><BR>[arXiv:1510.03823][INSPIRE]. 

[81ȟ Y.L. Dokshitzer. G.D. Leder. S. Morettk and B.R. Webber. Better jet clustering algorithms. JHEP 08 (1997˫ <BR>001 <BR>[hep-ph/9707323][INSPIRE]. 
[82ȟ S. Catank ev al.. Longitudinally invariant k⊥ clustering algorithms for hadron hadron collisions. Nucl. Phys. B 406 (1993˫ <BR>187. <BR>
√ 
[83ȟ 	ATLAS collaboration. Jet mass and substructure of inclusive jets in s 7 TeV pp collisions with the ATLAS experiment. <BR>JHEP 05 (2012˫ <BR>12: <BR>[arXiv:1203.4606][INSPIRE]. 
[84ȟ ATLAU collaboration. Performance of Missing Transverse Momentum Reconstruction in 
√ 
ATLAS with 2011 Proton-Proton Collisions at s 7 TeV. ATLAS-CONF-2012-101 <BR>(2012). 
JHEP06(2016)093 

√ 
[85ȟ ATLAU collaboration. Improved luminosity determination in pp collisions at s 7 TeV using the ATLAS detector at the LHC. Eur. Phys. J. C 73 (2013˫ <BR>251: <BR>[arXiv:1302.4393ȟ <BR>[INSPIRE]. 
[86ȟ 	ATLAS collaboration. Performance of b-Jet Identification in the ATLAS Experiment. 2018 <BR>JINST 11 P0400: <BR>[arXiv:1512.01094][INSPIRE]. 
[87ȟ J.M. Butterworth. B.E. Coz and J.R. Forshaw. WW scattering at the CERN LHC. Phys. Rev. D 65 (2002˫ <BR>096014 <BR>[hep-ph/0201098][INSPIRE]. 
[88ȟ 	ATLAS collaboration. Identification of boosted, hadronically decaying W bosons and 
√ 
comparisons with ATLAS data taken at s : TeV. Eur. Phys. J. C 76 (2016˫ <BR>154 <BR>[arXiv:1510.05821][INSPIRE]. 
[89ȟ J. Thalet and K. Vap Tilburg. Identifying Boosted Objects with N-subjettiness. JHEP 03 (2011˫ <BR>017 <BR>[arXiv:1011.2268][INSPIRE]. 
[90ȟ J. Thalet and K. Vap Tilburg. Maximizing Boosted Top Identification by Minimizing N-subjettiness. <BR>JHEP 02 (2012˫ <BR>095 <BR>[arXiv:1108.2701][INSPIRE]. 
[91ȟ 	CMS collaboration. Identification techniques for highly boosted W bosons that decay into hadrons. <BR>JHEP 12 (2014˫ <BR>017 <BR>[arXiv:1410.4227][INSPIRE]. 
[92ȟ 	ATLAS collaboration. A new method to distinguish hadronically decaying boosted Z bosons from W bosons using the ATLAS detector. Eur. Phys. J. C 76 (2016˫ <BR>23: <BR>[arXiv:1509.04939][INSPIRE]. 
[93ȟ G. Kasieczka. T. Plehn. T. Schell. T. Streblet and G.P. Salam. Resonance Searches with an Updated Top Tagger. <BR>JHEP 06 (2015˫ <BR>205 <BR>[arXiv:1503.05921][INSPIRE]. 
[94ȟ ATLAU collaboration. Calibration of the performance of b-tagging for c and light-flavour jets in the 2012 ATLAS data. ATLAS-CONF-2014-048 <BR>(2014). 
[95ȟ ATLAU collaboration. Calibration of the performance of b-tagging for c and light-flavour jets in the 2012 ATLAS data. ATLAS-CONF-2014-048 <BR>(2014). 
[96ȟ S. Sch¨atzen and M. Spannowsky. Tagging highly boosted top quarks. Phys. Rev. D 89 (2014˫ <BR>014007 <BR>[arXiv:1308.0540][INSPIRE]. 
JHEP06(2016)093 

The ATLAS collaboration 

G. Aad86. B. Abbott114. J. Abdallah152. O. Abdinov11. R. Aben108. M. Abolins91 . O.S. AbouZeid159. H. Abramowicz154. H. Abreu153. R. Abreu117. Y. Abulaiti147a,147b . 
B.S. 
Acharya164a,164b ,a. L. Adamczyk39a. D.L. Adams26. J. Adelman109. S. Adomeit101 . 

T. 
Adye132. A.A. Affolder75. T. Agatonovic-Jovin13. J. Agricola55. J.A. Aguilar-Saavedra127a,127f . 


S.P. Ahlen23. F. Ahmadov66,b. G. Aielli134a,134b . H. Akerstedt147a,147b . T.P.A. ˚,
Akesson82 
A.V. 
Akimov97. G.L. Alberghi21a,21b . J. Albert169. S. Albrand56. M.J. Alconada Verzini72 . 

M. 
Aleksa31. I.N. Aleksandrov66. C. Alexa27b. G. Alexander154. T. Alexopoulos10 . 

M. 
Alhroob114. G. Alimonti92a. L. Alio86. J. Alison32. S.P. Alkire36. B.M.M. Allbrooke150 . 


P.P. 
Allport18. A. Aloisio105a,105b . A. Alonso37. F. Alonso72. C. Alpigiani139. A. Altheimer36 . ´

B. 
Alvare— Gonzalez31. D. Alvare— Piqueras167. M.G. Alviggi105a,105b . B.T. Amadio15 . 

K. 
Amako67. Y. Amaran Coutinho25a. C. Amelung24. D. Amidei90. S.P. Amot Dos Santos127a,127c . 

A. 
Amorim127a,127b . S. Amoroso49. N. Amram154. G. Amundsen24. C. Anastopoulos140 . L.S. Ancu50. N. Andari109. T. Andeen36. C.F. Anders59b. G. Anders31. J.K. Anders75 . 


K.J. 
Anderson32. A. Andreazza92a,92b . V. Andrei59a. S. Angelidakis9. I. Angelozzi108. P. Anger45 . 

A. 
Angerami36. F. Anghinolfi31. A.V. Anisenkov110,c. N. Anjos12. A. Annovi125a,125b . 

M. 
Antonelli48. A. Antonov99. J. Antos145b . F. Anulli133a. M. Aoki67. L. Aperiq Bella18 . 

G. 
Arabidze91. Y. Arai67. J.P. Araque127a . A.T.H. Arce46. F.A. Arduh72. J-F. Arguin96 . 

S. 
Argyropoulos64. M. Arik19a. A.J. Armbruster31. O. Arnaez31. H. Arnold49. M. Arratia29 . O. Arslan22. A. Artamonov98. G. Artoni24. S. Asai156. N. Asbah43. A. Ashkenazi154 . 


B.˚,
Asman147a,147b . L. Asquith150. K. Assamagan26. R. Astalos145a . M. Atkinson166 
N.B. 
Atlay142. K. Augsten129. M. Aurousseau146b. G. Avolio31. B. Axen15. M.K. Ayoub118 . 

G. 
Azuelos96,d. M.A. Baak31. A.E. Baas59a. M.J. Baca18. C. Bacci135a,135b . H. Bachacou137 . 

K. 
Bachas155. M. Backes31. M. Backhaus31. P. Bagiacchi133a,133b . P. Bagnaia133a,133b . Y. Bai34a . 

T. 
Bain36. J.T. Baines132. O.K. Baker176. E.M. Baldin110,c. P. Balek130. T. Balestri149 . F. Balli85. W.K. Balunas123. E. Banas40. Sw. Banerjee173. A.A.E. Bannoura175. L. Barak31 . E.L. Barberio89. D. Barberis51a,51b . M. Barbero86. T. Barillari102. M. Barisonzi164a,164b . 

T. 
Barklow144. N. Barlow29. S.L. Barnes85. B.M. Barnett132. R.M. Barnett15. Z. Barnovska5 . 

A. 
Baroncelli135a. G. Barone24. A.J. Barr121. F. Barreiro83. J. Barreirq Guimar˜aes da Costa58 . R. Bartoldus144. A.E. Barton73. P. Bartos145a . A. Basalaev124. A. Bassalat118. A. Basye166 . R.L. Bates54. S.J. Batista159. J.R. Batley29. M. Battaglia138. M. Bauce133a,133b . F. Bauer137 . 


H.S. Bawa144,e. J.B. Beacham112. M.D. Beattie73. T. Beau81. P.H. Beauchemin162 . R. Beccherle125a,125b . P. Bechtle22. H.P. Beck17,f . K. Becker121. M. Becker84. M. Beckingham170 . C. Becot118. A.J. Beddall19b. A. Beddall19b. V.A. Bednyakov66. C.P. Bee149. L.J. Beemster108 . 
T.A. 
Beermann31. M. Begel26. J.K. Behr121. C. Belanger-Champagne88. W.H. Bell50. G. Bella154 . L. Bellagamba21a. A. Bellerive30. M. Bellomo87. K. Belotskiy99. O. Beltramello31. O. Benary154 . 

D. 
Benchekroun136a . M. Bender101. K. Bendtz147a,147b . N. Benekos10. Y. Benhammou154 . 

E. 
Benhat Noccioli50. J.A. Benite— Garcia160b . D.P. Benjamin46. J.R. Bensinger24 . 

S. 
Bentvelsen108. L. Beresford121. M. Beretta48. D. Berge108. E. Bergeaas Kuutmann165 . 

N. 
Berger5. F. Berghaus169. J. Beringer15. C. Bernard23. N.R. Bernard87. C. Bernius111 . F.U. Bernlochner22. T. Berry78. P. Berta130. C. Bertella84. G. Bertoli147a,147b . F. Bertolucci125a,125b . C. Bertsche114. D. Bertsche114. M.I. Besana92a. G.J. Besjes37 . O. Bessidskaia Bylund147a,147b . M. Bessner43. N. Besson137. C. Betancourt49. S. Bethke102 . 


A.J. 
Bevan77. W. Bhimji15. R.M. Bianchi126. L. Bianchini24. M. Bianco31. O. Biebel101 . 

D. 
Biedermann16. S.P. Bieniek79. N.V. Biesuz125a,125b . M. Biglietti135a . 

J. 
Bilbaq Dg Mendizabal50. H. Bilokon48. M. Bindi55. S. Binet118. A. Bingul19b. C. Bini133a,133b . 

S. 
Biondi21a,21b . D.M. Bjergaard46. C.W. Black151. J.E. Black144. K.M. Black23 . 


JHEP06(2016)093 

D. Blackburn139. R.E. Blair6. J.-B. Blanchard137. J.E. Blanco78. T. Blazek145a. I. Bloch43 . C. Blocker24. W. Blum84,∗. U. Blumenschein55. S. Blunier33a. G.J. Bobbink108 . 
V.S. 
Bobrovnikov110,c. S.S. Bocchetta82. A. Bocci46. C. Bock101. M. Boehler49. J.A. Bogaerts31 . 

D. 
Bogavac13. A.G. Bogdanchikov110. C. Bohm147a. V. Boisvert78. T. Bold39a. V. Boldea27b . 


A.S. 
Boldyrev100. M. Bomben81. M. Bona77. M. Boonekamp137. A. Borisov131. G. Borissov73 . 

S. 
Borroni43. J. Bortfeldt101. V. Bortolotto61a,61b,61c . K. Bos108. D. Boscherini21a. M. Bosman12 . 

J. 
Boudreau126. J. Bouffard2. E.V. Bouhova-Thacker73. D. Boumediene35. C. Bourdarios118 . 

N. 
Bousson115. S.K. Boutle54. A. Boveia31. J. Boyd31. I.R. Boyko66. I. Bozic13. J. Bracinik18 . 

A. 
Brandt8. G. Brandt55. O. Brandt59a. U. Bratzler157. B. Brau87. J.E. Brau117 . 


H.M. 
Braun175,∗. W.D. Breadep Madden54. K. Brendlinger123. A.J. Brennan89. L. Brenner108 . R. Brenner165. S. Bressler172. T.M. Bristow47. D. Britton54. D. Britzger43. F.M. Brochu29 . I. Brock22. R. Brock91. J. Bronner102. G. Brooijmans36. T. Brooks78. W.K. Brooks33b . 

J. 
Brosamer15. E. Brost117. P.A. Bruckmap dg Renstrom40. D. Bruncko145b. R. Bruneliere49 . 

A. 
Bruni21a. G. Bruni21a. M. Bruschi21a. N. Bruscino22. L. Bryngemark82. T. Buanes14 . 

Q. 
Buat143. P. Buchholz142. A.G. Buckley54. S.I. Buda27b. I.A. Budagov66. F. Buehrer49 . L. Bugge120. M.K. Bugge120. O. Bulekov99. D. Bullock8. H. Burckhart31. S. Burdin75 . C.D. Burgard49. B. Burghgrave109. S. Burke132. I. Burmeister44. E. Busato35. D. B¨uscher49 . 

V. 
B¨uscher84. P. Bussey54. J.M. Butler23. A.I. Butt3. C.M. Buttar54. J.M. Butterworth79 . 167

P. 
Butti108. W. Buttinger26. A. Buzatu54. A.R. Buzykaev110,c. S. Cabrera Urb´an. 

D. 
Caforio129. V.M. Cairo38a,38b . O. Cakir4a. N. Calace50. P. Calafiura15. A. Calandri137 . 

G. 
Calderini81. P. Calfayan101. L.P. Caloba25a. D. Calvet35. S. Calvet35. R. Camachq Toro32 . 

S. 
Camarda43. P. Camarri134a,134b . D. Cameron120. R. Caminan Armadans166. S. Campana31 . 

M. 
Campanelli79. A. Campoverde149. V. Canale105a,105b . A. Canepa160a. M. Canq Bret34e . 

J. 
Cantero83. R. Cantrill127a. T. Cao41. M.D.M. Capeans Garrido31. I. Caprini27b. M. Caprini27b . 

M. 
Capua38a,38b . R. Caputo84. R.M. Carbone36. R. Cardarelli134a. F. Cardillo49. T. Carli31 . 

G. 
Carlino105a. L. Carminati92a,92b . S. Caron107. E. Carquin33a. G.D. Carrillo-Montoya31 . J.R. Carter29. J. Carvalho127a,127c . D. Casadei79. M.P. Casado12,g. M. Casolino12 . 

E. 
Castaneda-Miranda146a. A. Castelli108. V. Castillq Gimenez167. N.F. Castro127a,h . 

P. 
Catastini58. A. Catinaccio31. J.R. Catmore120. A. Cattai31. J. Caudron84. V. Cavaliere166 . 

D. 
Cavalli92a. M. Cavalli-Sforza12. V. Cavasinni125a,125b . F. Ceradini135a,135b . B.C. Cerio46 . 

K. 
Cerny130. A.S. Cerqueira25b. A. Cerri150. L. Cerrito77. F. Cerutti15. M. Cerv31. A. Cervelli17 . 


S.A. Cetin19c. A. Chafaq136a. D. Chakraborty109. I. Chalupkova130. Y.L. Chan61a. P. Chang166 . 
J.D. 
Chapman29. D.G. Charlton18. C.C. Chau159. C.A. Chave— Barajas150. S. Cheatham153 . 

A. 
Chegwidden91. S. Chekanov6. S.V. Chekulaev160a. G.A. Chelkov66,i. M.A. Chelstowska90 . C. Chen65. H. Chen26. K. Chen149. L. Chen34d,j . S. Chen34c. S. Chen156. X. Chen34f. Y. Chen68 . H.C. Cheng90. Y. Cheng32. A. Cheplakov66. E. Cheremushkina131. R. Cherkaouk En Moursli136e . 

V. 
Chernyatin26,∗. E. Cheu7. L. Chevalier137. V. Chiarella48. G. Chiarelli125a,125b . 

G. 
Chiodini74a. A.S. Chisholm18. R.T. Chislett79. A. Chitan27b. M.V. Chizhov66. K. Choi62 . 

S. 
Chouridou9. B.K.B. Chow101. V. Christodoulou79. D. Chromek-Burckhart31. J. Chudoba128 . 


A.J. 
Chuinard88. J.J. Chwastowski40. L. Chytka116. G. Ciapetti133a,133b . A.K. Ciftci4a . 

D. 
Cinca54. V. Cindro76. I.A. Cioara22. A. Ciocio15. F. Cirotto105a,105b . Z.H. Citron172 . 

M. 
Ciubancan27b. A. Clark50. B.L. Clark58. P.J. Clark47. R.N. Clarke15. C. Clement147a,147b . Y. Coadou86. M. Cobal164a,164c . A. Coccaro50. J. Cochran65. L. Coffey24. J.G. Cogan144 . L. Colasurdo107. B. Cole36. S. Cole109. A.P. Colijn108. J. Collot56. T. Colombo59c . 

G. 
Compostella102. P. Condg Mui˜no127a,127b . E. Coniavitis49. S.H. Connell146b. I.A. Connelly78 . 

V. 
Consorti49. S. Constantinescu27b. C. Conta122a,122b . G. Conti31. F. Conventi105a,k . 

M. 
Cooke15. B.D. Cooper79. A.M. Cooper-Sarkar121. T. Cornelissen175. M. Corradi133a,133b . F. Corriveau88,l. A. Corso-Radu163. A. Cortes-Gonzalez12. G. Cortiana102. G. Costa92a . 


JHEP06(2016)093 

M.J. 
Costa167. D. Costanzo140. D. Cˆot´e8. G. Cottin29. G. Cowan78. B.E. Cox85. K. Cranmer111 . 

G. 
Cree30. S. Cr´ep´e-Renaudin56. F. Crescioli81. W.A. Cribbs147a,147b . M. Crispip Ortuzar121 . 

M. 
Cristinziani22. V. Croft107. G. Crosetti38a,38b . T. Cuhadat Donszelmann140. J. Cummings176 . 

M. 
Curatolo48. J. C´uth84. C. Cuthbert151. H. Czirr142. P. Czodrowski3. S. D’Auria54 . 

M. 
D’Onofrio75. M.J. Da Cunha Sargedas Dg Sousa127a,127b . C. Da Via85. W. Dabrowski39a . 

A. 
Dafinca121. T. Dai90. O. Dale14. F. Dallaire96. C. Dallapiccola87. M. Dam37. J.R. Dandoy32 . 


N.P. 
Dang49. A.C. Daniells18. M. Danninger168. M. Danq Hoffmann137. V. Dao49. G. Darbo51a . 

S. 
Darmora8. J. Dassoulas3. A. Dattagupta62. W. Davey22. C. David169. T. Davidek130 . 

E. 
Davies121,m. M. Davies154. P. Davison79. Y. Davygora59a. E. Dawe89. I. Dawson140 . R.K. Daya-Ishmukhametova87. K. De8. R. dg Asmundis105a. A. Dg Benedetti114 . 

S. 
Dg Castro21a,21b . S. Dg Cecco81. N. Dg Groot107. P. dg Jong108. H. Dg la Torre83 . F. Dg Lorenzi65. D. Dg Pedis133a. A. Dg Salvo133a. U. Dg Sanctis150. A. Dg Santo150 . 


J.B. 
Dg Vivig Dg Regie118. W.J. Dearnaley73. R. Debbe26. C. Debenedetti138. D.V. Dedovich66 . I. Deigaard108. J. Den Peso83. T. Den Prete125a,125b . D. Delgove118. F. Deliot137 . C.M. Delitzsch50. M. Deliyergiyev76. A. Dell’Acqua31. L. Dell’Asta23. M. Dell’Orso125a,125b . 

M. 
Della Pietra105a,k. D. della Volpe50. M. Delmastro5. P.A. Delsart56. C. Deluca108 . 


D.A. DeMarco159. S. Demers176. M. Demichev66. A. Demilly81. S.P. Denisov131. D. Derendarz40 . 
J.E. 
Derkaoui136d. F. Derue81. P. Dervan75. K. Desch22. C. Deterre43. K. Dette44 . 
P.O. Deviveiros31. A. Dewhurst132. S. Dhaliwal24. A. Dk Ciaccio134a,134b . L. Dk Ciaccio5 . 


A. 
Dk Domenico133a,133b . C. Dk Donato133a,133b . A. Dk Girolamo31. B. Dk Girolamo31 . 

A. 
Dk Mattia153. B. Dk Micco135a,135b . R. Dk Nardo48. A. Dk Simone49. R. Dk Sipio159 . 

D. 
Dk Valentino30. C. Diaconu86. M. Diamond159. F.A. Dias47. M.A. Diaz33a. E.B. Diehl90 . 

J. 
Dietrich16. S. Diglio86. A. Dimitrievska13. J. Dingfelder22. P. Dita27b. S. Dita27b. F. Dittus31 . F. Djama86. T. Djobava52b. J.I. Djuvsland59a. M.A.B. dq Vale25c. D. Dobos31. M. Dobre27b . C. Doglioni82. T. Dohmae156. J. Dolejsi130. Z. Dolezal130. B.A. Dolgoshein99,∗. M. Donadelli25d . 

S. 
Donati125a,125b . P. Dondero122a,122b . J. Donini35. J. Dopke132. A. Doria105a. M.T. Dova72 . 


A.T. 
Doyle54. E. Drechsler55. M. Dris10. E. Dubreuil35. E. Duchovni172. G. Duckeck101 . O.A. Ducu27b. D. Duda108. A. Dudarev31. L. Duflot118. L. Duguid78. M. D¨uhrssen31 . 

M. 
Dunford59a. H. Durap Yildiz4a. M. D¨uren53. A. Durglishvili52b. D. Duschinger45. B. Dutta43 . 

M. 
Dyndal39a. C. Eckardt43. K.M. Ecker102. R.C. Edgar90. W. Edson2. N.C. Edwards47 . 

W. 
Ehrenfeld22. T. Eifert31. G. Eigen14. K. Einsweiler15. T. Ekelof165. M. En Kacimi136c . 

M. 
Ellert165. S. Elles5. F. Ellinghaus175. A.A. Elliot169. N. Ellis31. J. Elmsheuser101. M. Elsing31 . 

D. 
Emeliyanov132. Y. Enari156. O.C. Endner84. M. Endo119. J. Erdmann44. A. Ereditato17 . 

G. 
Ernis175. J. Ernst2. M. Ernst26. S. Errede166. E. Ertel84. M. Escalier118. H. Esch44 . C. Escobar126. B. Esposito48. A.I. Etienvre137. E. Etzion154. H. Evans62. A. Ezhilov124 . L. Fabbri21a,21b . G. Facini32. R.M. Fakhrutdinov131. S. Falciano133a. R.J. Falla79. J. Faltova130 . Y. Fang34a. M. Fanti92a,92b . A. Farbin8. A. Farilla135a . T. Farooque12. S. Farrell15 . 


S.M. 
Farrington170. P. Farthouat31. F. Fassi136e. P. Fassnacht31. D. Fassouliotis9 . 

M. 
Faucck Giannelli78. A. Favareto51a,51b . L. Fayard118. O.L. Fedin124,n. W. Fedorko168 . 

S. 
Feigl31. L. Feligioni86. C. Feng34d. E.J. Feng31. H. Feng90. A.B. Fenyuk131. L. Feremenga8 . 

P. 
Fernande— Martinez167. S. Fernande— Perez31. J. Ferrando54. A. Ferrari165. P. Ferrari108 . R. Ferrari122a. D.E. Ferreira dg Lima54. A. Ferrer167. D. Ferrere50. C. Ferretti90 . 

A. 
Ferrettq Parodi51a,51b . M. Fiascaris32. F. Fiedler84. A. Filipˇciˇc76. M. Filipuzzi43 . F. Filthaut107. M. Fincke-Keeler169. K.D. Finelli151. M.C.N. Fiolhais127a,127c . L. Fiorini167 . 

A. 
Firan41. A. Fischer2. C. Fischer12. J. Fischer175. W.C. Fisher91. N. Flaschel43. I. Fleck142 . 

P. 
Fleischmann90. G.T. Fletcher140. G. Fletcher77. R.R.M. Fletcher123. T. Flick175. A. Floderus82 . L.R. Flores Castillo61a. M.J. Flowerdew102. A. Formica137. A. Forti85. D. Fournier118. H. Fox73 . 

S. 
Fracchia12. P. Francavilla81. M. Franchini21a,21b . D. Francis31. L. Franconi120. M. Franklin58 . 


JHEP06(2016)093 

M. Frate163. M. Fraternali122a,122b . D. Freeborn79. S.T. French29. F. Friedrich45 . 
D. Froidevaux31. J.A. Frost121. C. Fukunaga157. E. Fullana Torregrosa84. B.G. Fulsom144 . 
T. Fusayasu103. J. Fuster167. C. Gabaldon56. O. Gabizon175. A. Gabrielli21a,21b . A. Gabrielli15 . 
G.P. 
Gach18. S. Gadatsch31. S. Gadomski50. G. Gagliardi51a,51b . P. Gagnon62. C. Galea107 . 

B. 
Galhardo127a,127c . E.J. Gallas121. B.J. Gallop132. P. Gallus129. G. Galster37. K.K. Gan112 . ıa167

J. 
Gao34b,86 . Y. Gao47. Y.S. Gao144,e. F.M. Garay Walls47. F. Garberson176. C. Garc´. 


J.E. 
Garc´ıa Navarro167. M. Garcia-Sciveres15. R.W. Gardner32. N. Garelli144. V. Garonne120 . C. Gatti48. A. Gaudiello51a,51b . G. Gaudio122a. B. Gaur142. L. Gauthier96. P. Gauzzi133a,133b . I.L. Gavrilenko97. C. Gay168. G. Gaycken22. E.N. Gazis10. P. Ge34d. Z. Gecse168. C.N.P. Gee132 . Ch. Geich-Gimbel22. M.P. Geisler59a. C. Gemme51a. M.H. Genest56. S. Gentile133a,133b . 

M. 
George55. S. George78. D. Gerbaudo163. A. Gershon154. S. Ghasemi142. H. Ghazlane136b . 

B. 
Giacobbe21a. S. Giagu133a,133b . V. Giangiobbe12. P. Giannetti125a,125b . B. Gibbard26 . 


S.M. Gibson78. M. Gignac168. M. Gilchriese15. T.P.S. Gillam29. D. Gillberg31. G. Gilles35 . 
D.M. 
Gingrich3,d. N. Giokaris9. M.P. Giordani164a,164c . F.M. Giorgi21a. F.M. Giorgi16 . 
P.F. Giraud137. P. Giromini48. D. Giugni92a. C. Giuliani102. M. Giulini59b. B.K. Gjelsten120 . 


S. 
Gkaitatzis155. I. Gkialas155. E.L. Gkougkousis118. L.K. Gladilin100. C. Glasman83 . 

J. 
Glatzer31. P.C.F. Glaysher47. A. Glazov43. M. Goblirsch-Kolb102. J.R. Goddard77 . 

J. 
Godlewski40. S. Goldfarb90. T. Golling50. D. Golubkov131. A. Gomes127a,127b,127d 


. R. Gon¸calo127a. J. Goncalves Pintq Firminq Da Costa137. L. Gonella22. S. Gonz´ale— dg la Hoz167 . 
G. Gonzale— Parra12. S. Gonzalez-Sevilla50. L. Goossens31. P.A. Gorbounov98. H.A. Gordon26 . I. Gorelov106. B. Gorini31. E. Gorini74a,74b . A. Goriˇsek76. E. Gornicki40. A.T. Goshaw46 . C. G¨ossling44. M.I. Gostkin66. D. Goujdami136c. A.G. Goussiou139. N. Govender146b . 
E. Gozani153. H.M.X. Grabas138. L. Graber55. I. Grabowska-Bold39a. P.O.J. Gradin165 . 
P. Grafstr¨om21a,21b . K-J. Grahn43. J. Gramling50. E. Gramstad120. S. Grancagnolo16 . 
V. Gratchev124. H.M. Gray31. E. Graziani135a. Z.D. Greenwood80,o. C. Grefe22. K. Gregersen79 . 
I.M. Gregor43. P. Grenier144. J. Griffiths8. A.A. Grillo138. K. Grimm73. S. Grinstein12,p. 
Ph. Gris35. J.-F. Grivaz118. J.P. Grohs45. A. Grohsjean43. E. Gross172. J. Grosse-Knetter55 . 
G.C. Grossi80. Z.J. Grout150. L. Guan90. J. Guenther129. F. Guescini50. D. Guest163 . 
O. Gueta154. E. Guido51a,51b . T. Guillemin118. S. Guindon2. U. Gul54. C. Gumpert45. J. Guo34e . 
Y. Guo34b,q . S. Gupta121. G. Gustavino133a,133b . P. Gutierrez114. N.G. Gutierre— Ortiz79 . 
C. Gutschow45. C. Guyot137. C. Gwenlan121. C.B. Gwilliam75. A. Haas111. C. Haber15 . 

H.K. 
Hadavand8. N. Haddad136e. P. Haefner22. S. Hageb¨ock22. Z. Hajduk40. H. Hakobyan177 . 

M. 
Haleem43. J. Haley115. D. Hall121. G. Halladjian91. G.D. Hallewell86. K. Hamacher175 . 

P. 
Hamal116. K. Hamano169. A. Hamilton146a . G.N. Hamity140. P.G. Hamnett43. L. Han34b . 

K. 
Hanagaki67,r. K. Hanawa156. M. Hance138. B. Haney123. P. Hanke59a. R. Hanna137 . 


J.B. 
Hansen37. J.D. Hansen37. M.C. Hansen22. P.H. Hansen37. K. Hara161. A.S. Hard173 . 

T. 
Harenberg175. F. Hariri118. S. Harkusha93. R.D. Harrington47. P.F. Harrison170. F. Hartjes108 . 

M. 
Hasegawa68. Y. Hasegawa141. A. Hasib114. S. Hassani137. S. Haug17. R. Hauser91 . L. Hauswald45. M. Havranek128. C.M. Hawkes18. R.J. Hawkings31. A.D. Hawkins82 . 

T. 
Hayashi161. D. Hayden91. C.P. Hays121. J.M. Hays77. H.S. Hayward75. S.J. Haywood132 . 


S.J. 
Head18. T. Heck84. V. Hedberg82. L. Heelan8. S. Heim123. T. Heim175. B. Heinemann15 . L. Heinrich111. J. Hejbal128. L. Helary23. S. Hellman147a,147b . D. Hellmich22. C. Helsens12 . 

J. 
Henderson121. R.C.W. Henderson73. Y. Heng173. C. Hengler43. S. Henkelmann168 . 

A. 
Henrichs176. A.M. Henriques Correia31. S. Henrot-Versille118. G.H. Herbert16 . Y. Hern´ande— Jim´enez167. G. Herten49. R. Hertenberger101. L. Hervas31. G.G. Hesketh79 . 


N.P. Hessey108. J.W. Hetherly41. R. Hickling77. E. Hig´on-Rodriguez167. E. Hill169. J.C. Hill29 . 
K.H. 
Hiller43. S.J. Hillier18. I. Hinchliffe15. E. Hines123. R.R. Hinman15. M. Hirose158 . 

D. 
Hirschbuehl175. J. Hobbs149. N. Hod108. M.C. Hodgkinson140. P. Hodgson140. A. Hoecker31 . 


JHEP06(2016)093 

M.R. Hoeferkamp106. F. Hoenig101. M. Hohlfeld84. D. Hohn22. T.R. Holmes15. M. Homann44 . 
T.M. 
Hong126. W.H. Hopkins117. Y. Horii104. A.J. Horton143. J-Y. Hostachy56. S. Hou152 . 

A. 
Hoummada136a. J. Howard121. J. Howarth43. M. Hrabovsky116. I. Hristova16. J. Hrivnac118 . 

T. 
Hryn’ova5. A. Hrynevich94. C. Hsu146c. P.J. Hsu152,s. S.-C. Hsu139. D. Hu36. Q. Hu34b . X. Hu90. Y. Huang43. Z. Hubacek129. F. Hubaut86. F. Huegging22. T.B. Huffman121 . 


E.W. 
Hughes36. G. Hughes73. M. Huhtinen31. T.A. H¨ulsing84. N. Huseynov66,b. J. Huston91 . 

J. 
Huth58. G. Iacobucci50. G. Iakovidis26. I. Ibragimov142. L. Iconomidou-Fayard118. E. Ideal176 . Z. Idrissi136e . P. Iengo31. O. Igonkina108. T. Iizawa171. Y. Ikegami67. M. Ikeno67. Y. Ilchenko32,t . 

D. 
Iliadis155. N. Ilic144. T. Ince102. G. Introzzi122a,122b . P. Ioannou9,∗. M. Iodice135a . 

K. 
Iordanidou36. V. Ippolito58. A. Irles Quiles167. C. Isaksson165. M. Ishino69. M. Ishitsuka158 . R. Ishmukhametov112. C. Issever121. S. Istin19a. J.M. Iturbg Ponce85. R. Iuppa134a,134b . 

J. 
Ivarsson82. W. Iwanski40. H. Iwasaki67. J.M. Izen42. V. Izzo105a. S. Jabbar3. B. Jackson123 . 

M. 
Jackson75. P. Jackson1. M.R. Jaekel31. V. Jain2. K. Jakobs49. S. Jakobsen31. T. Jakoubek128 . 

J. 
Jakubek129. D.O. Jamin115. D.K. Jana80. E. Jansen79. R. Jansky63. J. Janssen22. M. Janus55 . 

G. 
Jarlskog82. N. Javadov66,b. T. Jav˚urek49. L. Jeanty15. J. Jejelava52a,u. G.-Y. Jeng151 . 

D. 
Jennens89. P. Jenni49,v . J. Jentzsch44. C. Jeske170. S. J´ez´equel5. H. Ji173. J. Jia149 . Y. Jiang34b. S. Jiggins79. J. Jimene— Pena167. S. Jin34a. A. Jinaru27b. O. Jinnouchi158 . 


M.D. 
Joergensen37. P. Johansson140. K.A. Johns7. W.J. Johnson139. K. Jon-And147a,147b . 

G. 
Jones170. R.W.L. Jones73. T.J. Jones75. J. Jongmanns59a. P.M. Jorge127a,127b . K.D. Joshi85 . 

J. 
Jovicevic160a . X. Ju173. P. Jussel63. A. Justg Rozas12,p. M. Kaci167. A. Kaczmarska40 . 

M. 
Kado118. H. Kagan112. M. Kagan144. S.J. Kahn86. E. Kajomovitz46. C.W. Kalderon121 . 

S. 
Kama41. A. Kamenshchikov131. N. Kanaya156. S. Kaneti29. V.A. Kantserov99. J. Kanzaki67 . 

B. 
Kaplan111. L.S. Kaplan173. A. Kapliy32. D. Kar146c. K. Karakostas10. A. Karamaoun3 . 

N. 
Karastathis10. M.J. Kareem55. E. Karentzos10. M. Karnevskiy84. S.N. Karpov66 . 
Z.M. Karpova66. K. Karthik111. V. Kartvelishvili73. A.N. Karyukhin131. K. Kasahara161 . 
L. Kashif173. R.D. Kass112. A. Kastanas14. Y. Kataoka156. C. Kato156. A. Katre50. J. Katzy43 . 


K. 
Kawade104. K. Kawagoe71. T. Kawamoto156. G. Kawamura55. S. Kazama156 . 
V.F. Kazanin110,c. R. Keeler169. R. Kehoe41. J.S. Keller43. J.J. Kempster78. H. Keoshkerian85 . 
O. Kepka128. B.P. Kerˇsevan76. S. Kersten175. R.A. Keyes88. F. Khalil-zada11 . 


H. 
Khandanyan147a,147b . A. Khanov115. A.G. Kharlamov110,c. T.J. Khoo29. V. Khovanskiy98 . 

E. 
Khramov66. J. Khubua52b,w. S. Kido68. H.Y. Kim8. S.H. Kim161. Y.K. Kim32. N. Kimura155 . O.M. Kind16. B.T. King75. M. King167. S.B. King168. J. Kirk132. A.E. Kiryunin102 . 

T. 
Kishimoto68. D. Kisielewska39a. F. Kiss49. K. Kiuchi161. O. Kivernyk137. E. Kladiva145b . 


M.H. 
Klein36. M. Klein75. U. Klein75. K. Kleinknecht84. P. Klimek147a,147b . A. Klimentov26 . R. Klingenberg44. J.A. Klinger140. T. Klioutchnikova31. E.-E. Kluge59a. P. Kluit108. S. Kluth102 . 

J. 
Knapik40. E. Kneringer63. E.B.F.G. Knoops86. A. Knue54. A. Kobayashi156. D. Kobayashi158 . 

T. 
Kobayashi156. M. Kobel45. M. Kocian144. P. Kodys130. T. Koffas30. E. Koffeman108 . L.A. Kogan121. S. Kohlmann175. Z. Kohout129. T. Kohriki67. T. Koi144. H. Kolanoski16 . 

M. 
Kolb59b. I. Koletsou5. A.A. Komar97,∗. Y. Komori156. T. Kondo67. N. Kondrashova43 . 

K. 
K¨oneke49. A.C. K¨onig107. T. Kono67,x. R. Konoplich111,y. N. Konstantinidis79 . R. Kopeliansky153. S. Koperny39a. L. K¨opke84. A.K. Kopp49. K. Korcyl40. K. Kordas155 . 

A. 
Korn79. A.A. Korol110,c. I. Korolkov12. E.V. Korolkova140. O. Kortner102. S. Kortner102 . 

T. 
Kosek130. V.V. Kostyukhin22. V.M. Kotov66. A. Kotwal46. A. Kourkoumeli-Charalampidi155 . C. Kourkoumelis9. V. Kouskoura26. A. Koutsman160a. R. Kowalewski169. T.Z. Kowalski39a . 

W. 
Kozanecki137. A.S. Kozhin131. V.A. Kramarenko100. G. Kramberger76. D. Krasnopevtsev99 . 


M.W. 
Krasny81. A. Krasznahorkay31. J.K. Kraus22. A. Kravchenko26. S. Kreiss111. M. Kretz59c . 

J. 
Kretzschmar75. K. Kreutzfeldt53. P. Krieger159. K. Krizka32. K. Kroeninger44. H. Kroha102 . 

J. 
Kroll123. J. Kroseberg22. J. Krstic13. U. Kruchonak66. H. Kr¨uger22. N. Krumnack65 . 


JHEP06(2016)093 

A. Kruse173. M.C. Kruse46. M. Kruskal23. T. Kubota89. H. Kucuk79. S. Kuday4b. S. Kuehn49 . 
A. Kugel59c. F. Kuger174. A. Kuhl138. T. Kuhl43. V. Kukhtin66. R. Kukla137. Y. Kulchitsky93 . 
S. Kuleshov33b. M. Kuna133a,133b . T. Kunigo69. A. Kupco128. H. Kurashige68. Y.A. Kurochkin93 . 
V. Kus128. E.S. Kuwertz169. M. Kuze158. J. Kvita116. T. Kwan169. D. Kyriazopoulos140 . 
A. La Rosa138. J.L. La Rosa Navarro25d. L. La Rotonda38a,38b . C. Lacasta167. F. Lacava133a,133b . 
J. Lacey30. H. Lacker16. D. Lacour81. V.R. Lacuesta167. E. Ladygin66. R. Lafaye5. B. Laforge81 . 
T. Lagouri176. S. Lai55. L. Lambourne79. S. Lammers62. C.L. Lampen7. W. Lampl7 . 
E. Lan¸con137. U. Landgraf49. M.P.J. Landon77. V.S. Lang59a. J.C. Lange12. A.J. Lankford163 . F. Lanni26. K. Lantzsch22. A. Lanza122a. S. Laplace81. C. Lapoire31. J.F. Laporte137. T. Lari92a . F. Lasagnk Manghi21a,21b . M. Lassnig31. P. Laurelli48. W. Lavrijsen15. A.T. Law138 . 
P. Laycock75. T. Lazovich58. O. Lg Dortz81. E. Lg Guirriec86. E. Lg Menedeu12. M. LeBlanc169 . 
T. LeCompte6. F. Ledroit-Guillon56. C.A. Lee146a . S.C. Lee152. L. Lee1. G. Lefebvre81 . 
M. Lefebvre169. F. Legger101. C. Leggett15. A. Lehan75. G. Lehmanp Miotto31. X. Lei7 . 
W.A. 
Leight30. A. Leisos155,z . A.G. Leister176. M.A.L. Leite25d. R. Leitner130. D. Lellouch172 . 

B. 
Lemmer55. K.J.C. Leney79. T. Lenz22. B. Lenzi31. R. Leone7. S. Leone125a,125b . 5


C. Leonidopoulos47. S. Leontsinis10. C. Leroy96. C.G. Lester29. M. Levchenko124. J. Levˆeque. 
D. Levin90. L.J. Levinson172. M. Levy18. A. Lewis121. A.M. Leyko22. M. Leyton42. B. Li34b,aa . 
H. Li149. H.L. Li32. L. Li46. L. Li34e. S. Li46. X. Li85. Y. Li34c,ab. Z. Liang138. H. Liao35 . 
B. Liberti134a. A. Liblong159. P. Lichard31. K. Lie166. J. Liebal22. W. Liebig14. C. Limbach22 . 
A. Limosani151. S.C. Lin152,ac. T.H. Lin84. F. Linde108. B.E. Lindquist149. J.T. Linnemann91 . 
E. Lipeles123. A. Lipniacka14. M. Lisovyi59b. T.M. Liss166. D. Lissauer26. A. Lister168 . 
A.M. 
Litke138. B. Liu152,ad. D. Liu152. H. Liu90. J. Liu86. J.B. Liu34b. K. Liu86. L. Liu166 . 

M. 
Liu46. M. Liu34b. Y. Liu34b. M. Livan122a,122b . A. Lleres56. J. Llorentg Merino83 . 
S.L. Lloyd77. F. Lq Sterzo152. E. Lobodzinska43. P. Loch7. W.S. Lockman138. F.K. Loebinger85 . 



A.E. 
Loevschall-Jensen37. K.M. Loew24. A. Loginov176. T. Lohse16. K. Lohwasser43 . 

M. 
Lokajicek128. B.A. Long23. J.D. Long166. R.E. Long73. K.A. Looper112. L. Lopes127a . 

D. 
Lope— Mateos58. B. Lope— Paredes140. I. Lope— Paz12. J. Lorenz101. N. Lorenzq Martinez62 . 

M. 
Losada20. P.J. L¨osel101. X. Lou34a. A. Lounis118. J. Love6. P.A. Love73. H. Lu61a. N. Lu90 . 


H.J. Lubatti139. C. Luci133a,133b . A. Lucotte56. C. Luedtke49. F. Luehring62. W. Lukas63 . 
L. Luminari133a. O. Lundberg147a,147b . B. Lund-Jensen148. D. Lynn26. R. Lysak128. E. Lytken82 . 76
H. Ma26. L.L. Ma34d. G. Maccarrone48. A. Macchiolo102. C.M. Macdonald140. B. Maˇcek. 
J. Machadq Miguens123,127b . D. Macina31. D. Madaffari86. R. Madar35. H.J. Maddocks73 . W.F. Mader45. A. Madsen165. J. Maeda68. S. Maeland14. T. Maeno26. A. Maevskiy100 . 
E. Magradze55. K. Mahboubi49. J. Mahlstedt108. C. Maiani137. C. Maidantchik25a . 
A.A. Maier102. T. Maier101. A. Maio127a,127b,127d . S. Majewski117. Y. Makida67. N. Makovec118 . 
B. Malaescu81. Pa. Malecki40. V.P. Maleev124. F. Malek56. U. Mallik64. D. Malon6. C. Malone144 . 
S. Maltezos10. V.M. Malyshev110. S. Malyukov31. J. Mamuzic43. G. Mancini48. B. Mandelli31 . L. Mandelli92a. I. Mandi´c76. R. Mandrysch64. J. Maneira127a,127b . A. Manfredini102 . L. Manhaes dg Andradg Filho25b. J. Manjarres Ramos160b. A. Mann101 . 
A. Manousakis-Katsikakis9. B. Mansoulie137. R. Mantifel88. M. Mantoani55. L. Mapelli31 . L. March146c. G. Marchiori81. M. Marcisovsky128. C.P. Marino169. M. Marjanovic13 . 
D.E. 
Marley90. F. Marroquim25a. S.P. Marsden85. Z. Marshall15. L.F. Marti17 . 

S. 
Marti-Garcia167. B. Martin91. T.A. Martin170. V.J. Martin47. B. Martip div Latour14 . 

M. 
Martinez12,p. S. Martin-Haugh132. V.S. Martoiu27b. A.C. Martyniuk79. M. Marx139 . F. Marzano133a. A. Marzin31. L. Masetti84. T. Mashimo156. R. Mashinistov97. J. Masik85 . A.L. Maslennikov110,c. I. Massa21a,21b . L. Massa21a,21b . P. Mastrandrea5 . 

A. 
Mastroberardino38a,38b . T. Masubuchi156. P. M¨attig175. J. Mattmann84. J. Maurer27b . 


S.J. Maxfield75. D.A. Maximov110,c. R. Mazini152. S.M. Mazza92a,92b . G. Me Goldrick159 . 
JHEP06(2016)093 

S.P. Me Kee90. A. McCarn90. R.L. McCarthy149. T.G. McCarthy30. N.A. McCubbin132 . 
K.W. 
McFarlane57,∗. J.A. Mcfayden79. G. Mchedlidze55. S.J. McMahon132. R.A. McPherson169,l . 

M. 
Medinnis43. S. Meehan146a. S. Mehlhase101. A. Mehta75. K. Meier59a. C. Meineck101 . 

B. 
Meirose42. B.R. Melladq Garcia146c. F. Meloni17. A. Mengarelli21a,21b . S. Menke102 . 

E. 
Meoni162. K.M. Mercurio58. S. Mergelmeyer22. P. Mermod50. L. Merola105a,105b . C. Meroni92a. F.S. Merritt32. A. Messina133a,133b . J. Metcalfe26. A.S. Mete163. C. Meyer84 . C. Meyer123. J-P. Meyer137. J. Meyer108. H. Meyet Zw Theenhausen59a. R.P. Middleton132 . 


76
S. Miglioranzi164a,164c . L. Mijovi´c22. G. Mikenberg172. M. Mikestikova128. M. Mikuˇz. 
M. Milesi89. A. Milic31. D.W. Miller32. C. Mills47. A. Milov172. D.A. Milstead147a,147b . 
A.A. 
Minaenko131. Y. Minami156. I.A. Minashvili66. A.I. Mincer111. B. Mindur39a. M. Mineev66 . Y. Ming173. L.M. Mir12. K.P. Mistry123. T. Mitani171. J. Mitrevski101. V.A. Mitsou167 . 

A. 
Miucci50. P.S. Miyagawa140. J.U. Mj¨ornmark82. T. Moa147a,147b . K. Mochizuki86 . 

S. 
Mohapatra36. W. Mohr49. S. Molander147a,147b . R. Moles-Valls22. R. Monden69. K. M¨onig43 . C. Monini56. J. Monk37. E. Monnier86. A. Montalbano149. J. Montejq Berlingen12 . F. Monticelli72. S. Monzani133a,133b . R.W. Moore3. N. Morange118. D. Moreno20 . 

M. 
Morenq Ll´acer55. P. Morettini51a. D. Mori143. T. Mori156. M. Morii58. M. Morinaga156 . 

V. 
Morisbak120. S. Moritz84. A.K. Morley151. G. Mornacchi31. J.D. Morris77. S.S. Mortensen37 . 

A. 
Morton54. L. Morvaj104. M. Mosidze52b. J. Moss144. K. Motohashi158. R. Mount144 . 

E. 
Mountricha26. S.V. Mouraviev97,∗. E.J.W. Moyse87. S. Muanza86. R.D. Mudd18 . F. Mueller102. J. Mueller126. R.S.P. Mueller101. T. Mueller29. D. Muenstermann50. P. Mullen54 . 


G.A. Mullier17. J.A. Murillq Quijada18. W.J. Murray170,132 . H. Musheghyan55. E. Musto153 . 
A.G. 
Myagkov131,ae. M. Myska129. B.P. Nachman144. O. Nackenhorst55. J. Nadal55. K. Nagai121 . R. Nagai158. Y. Nagai86. K. Nagano67. A. Nagarkar112. Y. Nagasaka60. K. Nagata161 . 

M. 
Nagel102. E. Nagy86. A.M. Nairz31. Y. Nakahama31. K. Nakamura67. T. Nakamura156 . I. Nakano113. H. Namasivayam42. R.F. Naranjq Garcia43. R. Narayan32. D.I. Narrias Villar59a . 

T. 
Naumann43. G. Navarro20. R. Nayyar7. H.A. Neal90. P.Yu. Nechaeva97. T.J. Neep85 . 


P.D. 
Nef144. A. Negri122a,122b . M. Negrini21a. S. Nektarijevic107. C. Nellist118. A. Nelson163 . 

S. 
Nemecek128. P. Nemethy111. A.A. Nepomuceno25a. M. Nessi31,af . M.S. Neubauer166 . 

M. 
Neumann175. R.M. Neves111. P. Nevski26. P.R. Newman18. D.H. Nguyen6. R.B. Nickerson121 . R. Nicolaidou137. B. Nicquevert31. J. Nielsen138. N. Nikiforou36. A. Nikiforov16 . 

V. 
Nikolaenko131,ae. I. Nikolic-Audit81. K. Nikolopoulos18. J.K. Nilsen120. P. Nilsson26 . Y. Ninomiya156. A. Nisati133a . R. Nisius102. T. Nobe156. L. Nodulman6. M. Nomachi119 . I. Nomidis30. T. Nooney77. S. Norberg114. M. Nordberg31. O. Novgorodova45. S. Nowak102 . 

M. 
Nozaki67. L. Nozka116. K. Ntekas10. G. Nunes Hanninger89. T. Nunnemann101. E. Nurse79 . F. Nuti89. B.J. O’Brien47. F. O’grady7. D.C. O’Neil143. V. O’Shea54. F.G. Oakham30,d . 

H. 
Oberlack102. T. Obermann22. J. Ocariz81. A. Ochi68. I. Ochoa36. J.P. Ochoa-Ricoux33a . 

S. 
Oda71. S. Odaka67. H. Ogren62. A. Oh85. S.H. Oh46. C.C. Ohm15. H. Ohman165. H. Oide31 . 

W. 
Okamura119. H. Okawa161. Y. Okumura32. T. Okuyama67. A. Olariu27b. S.A. Olivares Pino47 . 

D. 
Oliveira Damazio26. A. Olszewski40. J. Olszowska40. A. Onofre127a,127e . K. Onogi104 . P.U.E. Onyisi32,t. C.J. Oram160a . M.J. Oreglia32. Y. Oren154. D. Orestano135a,135b . 

N. 
Orlando155. C. Oropeza Barrera54. R.S. Orr159. B. Osculati51a,51b . R. Ospanov85 . 

G. 
Oterq y Garzon28. H. Otono71. M. Ouchrif136d. F. Ould-Saada120. A. Ouraou137 . 


K.P. 
Oussoren108. Q. Ouyang34a. A. Ovcharova15. M. Owen54. R.E. Owen18. V.E. Ozcan19a . 49

N. 
Ozturk8. K. Pachal143. A. Pachecq Pages12. C. Padilla Aranda12. M. Pag´aˇcov´a. 

S. 
Pagap Griso15. E. Paganis140. F. Paige26. P. Pais87. K. Pajchel120. G. Palacino160b . 


S. 
Palestini31. M. Palka39b. D. Pallin35. A. Palma127a,127b . Y.B. Pan173 . 
E.St. Panagiotopoulou10. C.E. Pandini81. J.G. Pandurq Vazquez78. P. Pani147a,147b . 


S. 
Panitkin26. D. Pantea27b. L. Paolozzi50. Th.D. Papadopoulou10. K. Papageorgiou155 . 



JHEP06(2016)093 

A. Paramonov6. D. Paredes Hernandez155. M.A. Parker29. K.A. Parker140. F. Parodi51a,51b . 
J.A. 
Parsons36. U. Parzefall49. E. Pasqualucci133a. S. Passaggio51a. F. Pastore135a,135b ,∗ . Fr. Pastore78. G. P´asztor30. S. Pataraia175. N.D. Patel151. J.R. Pater85. T. Pauly31. J. Pearce169 . 

B. 
Pearson114. L.E. Pedersen37. M. Pedersen120. S. Pedraza Lopez167. R. Pedro127a,127b . 


S.V. 
Peleganchuk110,c. D. Pelikan165. O. Penc128. C. Peng34a. H. Peng34b. B. Penning32 . 167

J. 
Penwell62. D.V. Perepelitsa26. E. Pere— Codina160a . M.T. P´ere— Garc´ıa-Esta˜n. L. Perini92a,92b . H. Pernegger31. S. Perrella105a,105b . R. Peschke43. V.D. Peshekhonov66 . 

K. 
Peters31. R.F.Y. Peters85. B.A. Petersen31. T.C. Petersen37. E. Petit43. A. Petridis1 . C. Petridou155. P. Petroff118. E. Petrolo133a. F. Petrucci135a,135b . N.E. Pettersson158 . R. Pezoa33b. P.W. Phillips132. G. Piacquadio144. E. Pianori170. A. Picazio50. E. Piccaro77 . 

M. 
Piccinini21a,21b . M.A. Pickering121. R. Piegaia28. D.T. Pignotti112. J.E. Pilcher32 . 


A.D. 
Pilkington85. A.W.J. Pin85. J. Pina127a,127b,127d . M. Pinamonti164a,164c ,ag. J.L. Pinfold3 . 

A. 
Pingel37. S. Pires81. H. Pirumov43. M. Pitt172. C. Pizio92a,92b . L. Plazak145a. M.-A. Pleier26 . 

V. 
Pleskot130. E. Plotnikova66. P. Plucinski147a,147b . D. Pluth65. R. Poettgen147a,147b . L. Poggioli118. D. Pohl22. G. Polesello122a. A. Poley43. A. Policicchio38a,38b . R. Polifka159 . 

A. 
Polini21a. C.S. Pollard54. V. Polychronakos26. K. Pomm`es31. L. Pontecorvo133a. B.G. Pope91 . 


G.A. 
Popeneciu27c. D.S. Popovic13. A. Poppleton31. S. Pospisil129. K. Potamianos15 . I.N. Potrap66. C.J. Potter150. C.T. Potter117. G. Poulard31. J. Poveda31. V. Pozdnyakov66 . 

P. 
Pralavorio86. A. Pranko15. S. Prasad31. S. Prell65. D. Price85. L.E. Price6. M. Primavera74a . 

S. 
Prince88. M. Proissl47. K. Prokofiev61c. F. Prokoshin33b. E. Protopapadaki137 . 

S. 
Protopopescu26. J. Proudfoot6. M. Przybycien39a. E. Ptacek117. D. Puddu135a,135b . 

E. 
Pueschel87. D. Puldon149. M. Purohit26,ah. P. Puzo118. J. Qian90. G. Qin54. Y. Qin85 . 

A. 
Quadt55. D.R. Quarrie15. W.B. Quayle164a,164b . M. Queitsch-Maitland85. D. Quilty54 . 

S. 
Raddum120. V. Radeka26. V. Radescu43. S.K. Radhakrishnan149. P. Radloff117. P. Rados89 . F. Ragusa92a,92b . G. Rahal178. S. Rajagopalan26. M. Rammensee31. C. Rangel-Smith165 . F. Rauscher101. S. Rave84. T. Ravenscroft54. M. Raymond31. A.L. Read120. N.P. Readioff75 . 


D.M. 
Rebuzzi122a,122b . A. Redelbach174. G. Redlinger26. R. Reece138. K. Reeves42 . L. Rehnisch16. J. Reichert123. H. Reisin28. C. Rembser31. H. Ren34a. A. Renaud118 . 

M. 
Rescigno133a . S. Resconi92a. O.L. Rezanova110,c. P. Reznicek130. R. Rezvani96. R. Richter102 . 

S. 
Richter79. E. Richter-Was39b. O. Ricken22. M. Ridel81. P. Rieck16. C.J. Riegel175. J. Rieger55 . O. Rifki114. M. Rijssenbeek149. A. Rimoldi122a,122b . L. Rinaldi21a. B. Risti´c50. E. Ritsch31 . I. Riu12. F. Rizatdinova115. E. Rizvi77. S.H. Robertson88,l. A. Robichaud-Veronneau88 . 

D. 
Robinson29. J.E.M. Robinson43. A. Robson54. C. Roda125a,125b . S. Roe31. O. Røhne120 . 

S. 
Rolli162. A. Romaniouk99. M. Romano21a,21b . S.M. Romanq Saez35. E. Romerq Adam167 . 

N. 
Rompotis139. M. Ronzani49. L. Roos81. E. Ros167. S. Rosati133a. K. Rosbach49. P. Rose138 . 
P.L. Rosendahl14. O. Rosenthal142. V. Rossetti147a,147b . E. Rossi105a,105b . L.P. Rossi51a . 



J.H.N. Rosten29. R. Rosten139. M. Rotaru27b. I. Roth172. J. Rothberg139. D. Rousseau118 . 
C.R. Royon137. A. Rozanov86. Y. Rozen153. X. Ruan146c. F. Rubbo144. I. Rubinskiy43 . 
V.I. Rud100. C. Rudolph45. M.S. Rudolph159. F. R¨uhr49. A. Ruiz-Martinez31. Z. Rurikova49 . 

N.A. 
Rusakovich66. A. Ruschke101. H.L. Russell139. J.P. Rutherfoord7. N. Ruthmann31 . 
Y.F. Ryabov124. M. Rybar166. G. Rybkin118. N.C. Ryder121. A. Ryzhov131. A.F. Saavedra151 . 


G. 
Sabato108. S. Sacerdoti28. A. Saddique3. H.F-W. Sadrozinski138. R. Sadykov66 . 
F. Safak Tehrani133a. P. Saha109. M. Sahinsoy59a. M. Saimpert137. T. Saito156. H. Sakamoto156 . 
Y. Sakurai171. G. Salamanna135a,135b . A. Salamon134a . J.E. Salazat Loyola33b. M. Saleem114 . 


D. 
Salek108. P.H. Sales Dg Bruin139. D. Salihagic102. A. Salnikov144. J. Salt167 . 

D. 
Salvatore38a,38b . F. Salvatore150. A. Salvucci61a. A. Salzburger31. D. Sammel49 . 

D. 
Sampsonidis155. A. Sanchez105a,105b . J. S´anchez167. V. Sanche— Martinez167. H. Sandaker120 . R.L. Sandbach77. H.G. Sander84. M.P. Sanders101. M. Sandhoff175. C. Sandoval20 . 


JHEP06(2016)093 

R. Sandstroem102. D.P.C. Sankey132. M. Sannino51a,51b . A. Sansoni48. C. Santoni35 . R. Santonico134a,134b . H. Santos127a. I. Santoyq Castillo150. K. Sapp126. A. Sapronov66 . 
J.G. 
Saraiva127a,127d . B. Sarrazin22. O. Sasaki67. Y. Sasaki156. K. Sato161. G. Sauvage5,∗ . 

E. 
Sauvan5. G. Savage78. P. Savard159,d. C. Sawyer132. L. Sawyer80,o. J. Saxon32. C. Sbarra21a . 

A. 
Sbrizzi21a,21b . T. Scanlon79. D.A. Scannicchio163. M. Scarcella151. V. Scarfone38a,38b . 

J. 
Schaarschmidt172. P. Schacht102. D. Schaefer31. R. Schaefer43. J. Schaeffer84. S. Schaepe22 . 

S. 
Schaetzel59b. U. Sch¨afer84. A.C. Schaffer118. D. Schaile101. R.D. Schamberger149. V. Scharf59a . 


V.A. 
Schegelsky124. D. Scheirich130. M. Schernau163. C. Schiavi51a,51b . C. Schillo49 . 

M. 
Schioppa38a,38b . S. Schlenker31. K. Schmieden31. C. Schmitt84. S. Schmitt59b. S. Schmitt43 . 

B. 
Schneider160a. Y.J. Schnellbach75. U. Schnoor45. L. Schoeffel137. A. Schoening59b . 


B.D. 
Schoenrock91. E. Schopf22. A.L.S. Schorlemmer55. M. Schott84. D. Schouten160a . 

J. 
Schovancova8. S. Schramm50. M. Schreyer174. N. Schuh84. M.J. Schultens22 . 
H.-C. Schultz-Coulon59a. H. Schulz16. M. Schumacher49. B.A. Schumm138. Ph. Schune137 . 
C. Schwanenberger85. A. Schwartzman144. T.A. Schwarz90. Ph. Schwegler102. H. Schweiger85 . 
Ph. Schwemling137. R. Schwienhorst91. J. Schwindling137. T. Schwindt22. F.G. Sciacca17 . 


E. 
Scifo118. G. Sciolla24. F. Scuri125a,125b . F. Scutti22. J. Searcy90. G. Sedov43. E. Sedykh124 . 

P. 
Seema22. S.C. Seidel106. A. Seiden138. F. Seifert129. J.M. Seixas25a. G. Sekhniaidze105a . 

K. 
Sekhon90. S.J. Sekula41. D.M. Seliverstov124,∗. N. Semprini-Cesari21a,21b . C. Serfon31 . L. Serin118. L. Serkin164a,164b . T. Serre86. M. Sessa135a,135b . R. Seuster160a. H. Severini114 . 

T. 
Sfiligoj76. F. Sforza31. A. Sfyrla31. E. Shabalina55. M. Shamim117. L.Y. Shan34a. R. Shang166 . 


J.T. 
Shank23. M. Shapiro15. P.B. Shatalov98. K. Shaw164a,164b . S.M. Shaw85 . 

A. 
Shcherbakova147a,147b . C.Y. Shehu150. P. Sherwood79. L. Shi152,ai. S. Shimizu68 . C.O. Shimmin163. M. Shimojima103. M. Shiyakova66,aj . A. Shmeleva97. D. Shoalej Saadi96 . 


M.J. 
Shochet32. S. Shojaii92a,92b . S. Shrestha112. E. Shulga99. M.A. Shupe7. S. Shushkevich43 . 

P. 
Sicho128. P.E. Sidebo148. O. Sidiropoulou174. D. Sidorov115. A. Sidoti21a,21b . F. Siegert45 . 
Dj. Sijacki13. J. Silva127a,127d . Y. Silver154. S.B. Silverstein147a. V. Simak129. O. Simard5 . 
Lj. Simic13. S. Simion118. E. Simioni84. B. Simmons79. D. Simon35. P. Sinervo159. N.B. Sinev117 . 



olin147a,147b
M. Sioli21a,21b . G. Siragusa174. A.N. Sisakyan66,∗. S.Yu. Sivoklokov100. J. Sj¨. 
T.B. 
Sjursen14. M.B. Skinner73. H.P. Skottowe58. P. Skubic114. M. Slater18. T. Slavicek129 . 

M. 
Slawinska108. K. Sliwa162. V. Smakhtin172. B.H. Smart47. L. Smestad14. S.Yu. Smirnov99 . Y. Smirnov99. L.N. Smirnova100,ak. O. Smirnova82. M.N.K. Smith36. R.W. Smith36 . 

M. 
Smizanska73. K. Smolek129. A.A. Snesarev97. G. Snidero77. S. Snyder26. R. Sobie169,l . F. Socher45. A. Soffer154. D.A. Soh152,ai. G. Sokhrannyi76. C.A. Solans Sanchez31. M. Solar129 . 

J. 
Solc129. E.Yu. Soldatov99. U. Soldevila167. A.A. Solodkov131. A. Soloshenko66 . O.V. Solovyanov131. V. Solovyev124. P. Sommer49. H.Y. Song34b,aa. N. Soni1. A. Sood15 . 

A. 
Sopczak129. B. Sopko129. V. Sopko129. V. Sorin12. D. Sosa59b. M. Sosebee8 . 


C.L. Sotiropoulou125a,125b . R. Soualah164a,164c . A.M. Soukharev110,c. D. South43. B.C. Sowden78 . 78
S. Spagnolo74a,74b . M. Spalla125a,125b . M. Spangenberg170. M. Spannowskyal. F. Span`o. W.R. Spearman58. D. Sperlich16. F. Spettel102. R. Spighi21a. G. Spigo31. L.A. Spiller89 . 
M. Spousta130. R.D. St. Denis54,∗. A. Stabile92a. S. Staerz45. J. Stahlman123. R. Stamen59a . 
S. Stamm16. E. Stanecka40. R.W. Stanek6. C. Stanescu135a. M. Stanescu-Bellu43 . 
M.M. Stanitzki43. S. Stapnes120. E.A. Starchenko131. J. Stark56. P. Staroba128. P. Starovoitov59a . R. Staszewski40. P. Steinberg26. B. Stelzer143. H.J. Stelzer31. O. Stelzer-Chilton160a. H. Stenzel53 . 
G.A. 
Stewart54. J.A. Stillings22. M.C. Stockton88. M. Stoebe88. G. Stoicea27b. P. Stolte55 . 

S. 
Stonjek102. A.R. Stradling8. A. Straessner45. M.E. Stramaglia17. J. Strandberg148 . 

S. 
Strandberg147a,147b . A. Strandlie120. E. Strauss144. M. Strauss114. P. Strizenec145b . R. Str¨ohmer174. D.M. Strom117. R. Stroynowski41. A. Strubig107. S.A. Stucci17. B. Stugu14 . 


N.A. Styles43. D. Su144. J. Su126. R. Subramaniam80. A. Succurro12. S. Suchek59a. Y. Sugaya119 . 
JHEP06(2016)093 

M. Suk129. V.V. Sulin97. S. Sultansoy4c. T. Sumida69. S. Sun58. X. Sun34a. J.E. Sundermann49 . 
K. Suruliz150. G. Susinno38a,38b . M.R. Sutton150. S. Suzuki67. M. Svatos128. M. Swiatlowski144 . I. Sykora145a. T. Sykora130. D. Ta49. C. Taccini135a,135b . K. Tackmann43. J. Taenzer159 . 
A. Taffard163. R. Tafirout160a . N. Taiblum154. H. Takai26. R. Takashima70. H. Takeda68 . 
T. Takeshita141. Y. Takubo67. M. Talby86. A.A. Talyshev110,c. J.Y.C. Tam174. K.G. Tan89 . 
J. Tanaka156. R. Tanaka118. S. Tanaka67. B.B. Tannenwald112. N. Tannoury22 . 
S. Tapia Araya33b. S. Tapprogge84. S. Tarem153. F. Tarrade30. G.F. Tartarelli92a. P. Tas130 . 
M. Tasevsky128. T. Tashiro69. E. Tassi38a,38b . A. Tavares Delgado127a,127b . Y. Tayalati136d . F.E. Taylor95. G.N. Taylor89. P.T.E. Taylor89. W. Taylor160b. F.A. Teischinger31 . 
P. Teixeira-Dias78. K.K. Temming49. D. Temple143. H. Tep Kate31. P.K. Teng152. J.J. Teoh119 . F. Tepel175. S. Terada67. K. Terashi156. J. Terron83. S. Terzo102. M. Testa48. R.J. Teuscher159,l . 
T. Theveneaux-Pelzer35. J.P. Thomas18. J. Thomas-Wilsker78. E.N. Thompson36 . 
P.D. 
Thompson18. R.J. Thompson85. A.S. Thompson54. L.A. Thomsen176. E. Thomson123 . 

M. 
Thomson29. R.P. Thun90,∗. M.J. Tibbetts15. R.E. Ticsg Torres86. V.O. Tikhomirov97,am . 


Yu.A. 
Tikhonov110,c. S. Timoshenko99. E. Tiouchichine86. P. Tipton176. S. Tisserant86 . 

K. 
Todome158. T. Todorov5,∗. S. Todorova-Nova130. J. Tojo71. S. Tok´ar145a. K. Tokushuku67 . 

K. 
Tollefson91. E. Tolley58. L. Tomlinson85. M. Tomoto104. L. Tompkins144,an. K. Toms106 . 

E. 
Torrence117. H. Torres143. E. Torr´q Pastor139. J. Toth86,ao. F. Touchard86. D.R. Tovey140 . 

T. 
Trefzger174. L. Tremblet31. A. Tricoli31. I.M. Trigger160a. S. Trincaz-Duvoid81 . 
M.F. Tripiana12. W. Trischuk159. B. Trocm´e56. C. Troncon92a. M. Trottier-McDonald15 . 


M. 
Trovatelli169. L. Truong164a,164c . M. Trzebinski40. A. Trzupek40. C. Tsarouchas31 . 
J.C-L. Tseng121. P.V. Tsiareshka93. D. Tsionou155. G. Tsipolitis10. N. Tsirintanis9 . 


S. 
Tsiskaridze12. V. Tsiskaridze49. E.G. Tskhadadze52a. K.M. Tsui61a. I.I. Tsukerman98 . 


V. 
Tsulaia15. S. Tsuno67. D. Tsybychev149. A. Tudorache27b. V. Tudorache27b. A.N. Tuna58 . 


S.A. Tupputi21a,21b . S. Turchikhin100,ak. D. Turecek129. R. Turra92a,92b . A.J. Turvey41 . 
P.M. 
Tuts36. A. Tykhonov50. M. Tylmad147a,147b . M. Tyndel132. I. Ueda156. R. Ueno30 . 

M. 
Ughetto147a,147b . M. Ugland14. F. Ukegawa161. G. Unal31. A. Undrus26. G. Unel163 . F.C. Ungaro49. Y. Unno67. C. Unverdorben101. J. Urban145b. P. Urquijo89. P. Urrejola84 . 

G. 
Usai8. A. Usanova63. L. Vacavant86. V. Vacek129. B. Vachon88. C. Valderanis84 . 

N. 
Valencic108. S. Valentinetti21a,21b . A. Valero167. L. Valery12. S. Valkar130. S. Vallecorsa50 . 


J.A. 
Valls Ferrer167. W. Vap Dep Wollenberg108. P.C. Vap Det Deijl108. R. vap det Geer108 . 

H. 
vap det Graaf108. N. vap Eldik153. P. vap Gemmeren6. J. Vap Nieuwkoop143. I. vap Vulpen108 . M.C. vap Woerden31. M. Vanadia133a,133b . W. Vandelli31. R. Vanguri123. A. Vaniachine6 . F. Vannucci81. G. Vardanyan177. R. Vari133a. E.W. Varnes7. T. Varol41. D. Varouchas81 . 

A. 
Vartapetian8. K.E. Varvell151. F. Vazeille35. T. Vazque— Schroeder88. J. Veatch7 . L.M. Veloce159. F. Veloso127a,127c . T. Velz22. S. Veneziano133a . A. Ventura74a,74b . D. Ventura87 . 

M. 
Venturi169. N. Venturi159. A. Venturini24. V. Vercesi122a. M. Verducci133a,133b . 

W. 
Verkerke108. J.C. Vermeulen108. A. Vest45,ap. M.C. Vetterli143,d. O. Viazlo82. I. Vichou166 . 

T. 
Vickey140. O.E. Vickey Boeriu140. G.H.A. Viehhauser121. S. Viel15. R. Vigne63 . 

M. 
Villa21a,21b . M. Villaplana Perez92a,92b . E. Vilucchi48. M.G. Vincter30. V.B. Vinogradov66 . I. Vivarelli150. F. Vives Vaque3. S. Vlachos10. D. Vladoiu101. M. Vlasak129. M. Vogel33a . 

P. 
Vokac129. G. Volpi125a,125b . M. Volpi89. H. vop det Schmitt102. H. vop Radziewski49 . 

E. 
vop Toerne22. V. Vorobel130. K. Vorobev99. M. Vos167. R. Voss31. J.H. Vossebeld75 . 

N. 
Vranjes13. M. Vranjes Milosavljevic13. V. Vrba128. M. Vreeswijk108. R. Vuillermet31 . I. Vukotic32. Z. Vykydal129. P. Wagner22. W. Wagner175. H. Wahlberg72. S. Wahrmund45 . 

J. 
Wakabayashi104. J. Walder73. R. Walker101. W. Walkowiak142. C. Wang152. F. Wang173 . 

H. 
Wang15. H. Wang41. J. Wang43. J. Wang151. K. Wang88. R. Wang6. S.M. Wang152 . 

T. 
Wang22. T. Wang36. X. Wang176. C. Wanotayaroj117. A. Warburton88. C.P. Ward29 . 


JHEP06(2016)093 

D.R. Wardrope79. A. Washbrook47. C. Wasicki43. P.M. Watkins18. A.T. Watson18 . I.J. Watson151. M.F. Watson18. G. Watts139. S. Watts85. B.M. Waugh79. S. Webb85 . 
M.S. 
Weber17. S.W. Weber174. J.S. Webster32. A.R. Weidberg121. B. Weinert62. J. Weingarten55 . C. Weiser49. H. Weits108. P.S. Wells31. T. Wenaus26. T. Wengler31. S. Wenig31. N. Wermes22 . 

M. 
Werner49. P. Werner31. M. Wessels59a. J. Wetter162. K. Whalen117. A.M. Wharton73 . 

A. 
White8. M.J. White1. R. White33b. S. White125a,125b . D. Whiteson163. F.J. Wickens132 . 

W. 
Wiedenmann173. M. Wielers132. P. Wienemann22. C. Wiglesworth37. L.A.M. Wiik-Fuchs22 . 

A. 
Wildauer102. H.G. Wilkens31. H.H. Williams123. S. Williams108. C. Willis91. S. Willocq87 . 

A. 
Wilson90. J.A. Wilson18. I. Wingerter-Seez5. F. Winklmeier117. B.T. Winter22. M. Wittgen144 . 

J. 
Wittkowski101. S.J. Wollstadt84. M.W. Wolter40. H. Wolters127a,127c . B.K. Wosiek40 . 

J. 
Wotschack31. M.J. Woudstra85. K.W. Wozniak40. M. Wu56. M. Wu32. S.L. Wu173. X. Wu50 . Y. Wu90. T.R. Wyatt85. B.M. Wynne47. S. Xella37. D. Xu34a. L. Xu26. B. Yabsley151 . 

S. 
Yacoob146a. R. Yakabe68. M. Yamada67. D. Yamaguchi158. Y. Yamaguchi119. A. Yamamoto67 . 

S. 
Yamamoto156. T. Yamanaka156. K. Yamauchi104. Y. Yamazaki68. Z. Yan23. H. Yang34e . 

H. 
Yang173. Y. Yang152. W-M. Yao15. Y.C. Yap81. Y. Yasu67. E. Yatsenko5. K.H. Yaw Wong22 . 

J. 
Ye41. S. Ye26. I. Yeletskikh66. A.L. Yen58. E. Yildirim43. K. Yorita171. R. Yoshida6 . 

K. 
Yoshihara123. C. Young144. C.J.S. Young31. S. Youssef23. D.R. Yu15. J. Yu8. J.M. Yu90 . 

J. 
Yu115. L. Yuan68. S.P.Y. Yuen22. A. Yurkewicz109. I. Yusuff29,aq . B. Zabinski40. R. Zaidan64 . 


A.M. Zaitsev131,ae. J. Zalieckas14. A. Zaman149. S. Zambito58. L. Zanello133a,133b . D. Zanzi89 . 145a
C. Zeitnitz175. M. Zeman129. A. Zemla39a. Q. Zeng144. K. Zengel24. O. Zenin131. T. Zeniˇˇs. 
D. Zerwas118. D. Zhang90. F. Zhang173. G. Zhang34b. H. Zhang34c. J. Zhang6. L. Zhang49 . R. Zhang34b,j . X. Zhang34d. Z. Zhang118. X. Zhao41. Y. Zhao34d,118 . Z. Zhao34b . 
A. Zhemchugov66. J. Zhong121. B. Zhou90. C. Zhou46. L. Zhou36. L. Zhou41. M. Zhou149 . 
N. Zhou34f. C.G. Zhu34d. H. Zhu34a. J. Zhu90. Y. Zhu34b. X. Zhuang34a. K. Zhukov97 . 
A. Zibell174. D. Zieminska62. N.I. Zimine66. C. Zimmermann84. S. Zimmermann49. Z. Zinonos55 . 
M. Zinser84. M. Ziolkowski142,L. ˇc,
Zivkovi´13. G. Zobernig173. A. Zoccoli21a,21b . M. zut Nedden16 
G. Zurzolo105a,105b and L. Zwalinski31 . 
1 Department of Physics, University of Adelaide, Adelaide, Australia 
2 Physics Department, SUNY Albany, Albany NY, U.S.A. 
3 Department of Physics, University of Alberta, Edmonton AB, Canada 
4(a) 
Department of Physics, Ankara University, Ankara; (b) Istanbul Aydin University, Istanbul; (c) 
Division of Physics, TOBB University of Economics and Technology, Ankara, Turkey 
5 LAPP, CNRS/IN2P3 and Universit´e Savoie Mont Blanc, Annecy-le-Vieux, France 
6 High Energy Physics Division, Argonne National Laboratory, Argonne IL, U.S.A. 
7 Department of Physics, University of Arizona, Tucson AZ, U.S.A. 
8 Department of Physics, The University of Texas at Arlington, Arlington TX, U.S.A. 
9 Physics Department, University of Athens, Athens, Greece 
10 Physics Department, National Technical University of Athens, Zografou, Greece 
11 Institute of Physics, Azerbaijan Academy of Sciences, Baku, Azerbaijan 
12 Institut de F´ısica d’Altes Energies (IFAE), The Barcelona Institute of Science and Technology, 
Barcelona, Spain, Spain 
13 Institute of Physics, University of Belgrade, Belgrade, Serbia 
14 Department for Physics and Technology, University of Bergen, Bergen, Norway 
15 Physics Division, Lawrence Berkeley National Laboratory and University of California, Berkeley 
CA, U.S.A. 
16 Department of Physics, Humboldt University, Berlin, Germany 
17 Albert Einstein Center for Fundamental Physics and Laboratory for High Energy Physics, 
University of Bern, Bern, Switzerland 
18 School of Physics and Astronomy, University of Birmingham, Birmingham, U.K. 
JHEP06(2016)093 

19 	(a) 
Department of Physics, Bogazici University, Istanbul; (b) Department of Physics Engineering, 
Gaziantep University, Gaziantep; (c) Department of Physics, Dogus University, Istanbul, Turkey 
20 	Centro de Investigaciones, Universidad Antonio Narino, Bogota, Colombia 
21 	(a) INFN Sezione di Bologna; (b) Dipartimento di Fisica e Astronomia, Universit`a di Bologna, Bologna, Italy 
22 	Physikalisches Institut, University of Bonn, Bonn, Germany 
23 	Department of Physics, Boston University, Boston MA, U.S.A. 
24 	Department of Physics, Brandeis University, Waltham MA, U.S.A. 
25 	(a) 
Universidade Federal do Rio De Janeiro COPPE/EE/IF, Rio de Janeiro; (b) Electrical Circuits Department,FederalUniversity of Juiz deFora(UFJF), Juiz deFora; (c) Federal University of Sao Joao del Rei (UFSJ), Sao Joao del Rei; (d) Instituto de Fisica,Universidade de Sao Paulo, Sao Paulo, Brazil 
26 Physics Department, Brookhaven National Laboratory, Upton NY, U.S.A. 27 (a) 
Transilvania University of Brasov, Brasov, Romania; (b) National Institute of Physics and Nuclear Engineering, Bucharest; (c) National Institute for Research and Development of Isotopic and Molecular Technologies, Physics Department, Cluj Napoca; (d) University Politehnica Bucharest, Bucharest; (e)] West University in Timisoara, Timisoara, Romania 
28 	Departamento de F´ısica, Universidad de Buenos Aires, Buenos Aires, Argentina 
29 	Cavendish Laboratory, University of Cambridge, Cambridge, U.K. 
30 	Department of Physics, Carleton University, Ottawa ON, Canada 
31 	CERN, Geneva, Switzerland 
32 	Enrico Fermi Institute, University of Chicago, Chicago IL, U.S.A. 
33 	(a) 
Departamento deF´ısica, Pontificia Universidad Cat´olica de Chile, Santiago; (b) Departamento 34]
de F´ısica, Universidad T´ecnica Federico Santa Mar´ıa, Valpara´ıso, Chile (a) Institute of High Energy Physics, Chinese Academy of Sciences, Beijing; (b) Department of Modern Physics, University of Science and Technology of China, Anhui; (c) Department of Physics, Nanjing University, Jiangsu; (d) School of Physics, Shandong University, Shandong; (e) Department of Physics and Astronomy, Shanghai Key Laboratory for Particle Physics and Cosmology, Shanghai JiaoTongUniversity, Shanghai; (also affiliated with PKU-CHEP); (f) Physics Department, Tsinghua University, Beijing 100084, China 
35 	Laboratoire de Physique Corpusculaire, Clermont Universit´e and Universit´e Blaise Pascal and CNRS/IN2P3, Clermont-Ferrand, France 
36 	Nevis Laboratory, Columbia University, Irvington NY, U.S.A. 
37 	Niels Bohr Institute, University of Copenhagen, Kobenhavn, Denmark 
38 	(a) INFN Gruppo Collegato di Cosenza, Laboratori Nazionali di Frascati; (b) Dipartimento di Fisica, Universit`a della Calabria, Rende, Italy 
39 	(a) AGH University of Science and Technology, Faculty of Physics and Applied Computer Science, Krakow; (b) Marian Smoluchowski Institute of Physics, Jagiellonian University, Krakow, Poland 
40 	Institute of Nuclear Physics Polish Academy of Sciences, Krakow, Poland 
41 	Physics Department, Southern MethodistUniversity, Dallas TX,U.S.A. 
42 	Physics Department, University ofTexas at Dallas,Richardson TX,U.S.A. 
43 	DESY, Hamburg and Zeuthen, Germany 
44 	Institut f¨ur Experimentelle Physik IV, Technische Universit¨at Dortmund, Dortmund, Germany 
45 	Institut f¨ur Kern-und Teilchenphysik, Technische Universit¨at Dresden, Dresden, Germany 
46 	Department of Physics, Duke University, Durham NC, U.S.A. 
47 	SUPA -School of Physics and Astronomy, University of Edinburgh, Edinburgh, U.K. 
48 	INFN Laboratori Nazionali di Frascati, Frascati, Italy 
49 	Fakult¨at f¨ur Mathematik und Physik, Albert-Ludwigs-Universit¨at, Freiburg, Germany 
50 	Section de Physique, Universit´e de Gen`eve, Geneva, Switzerland 
51 	(a) INFN Sezione di Genova; (b) Dipartimento di Fisica, Universit`a di Genova, Genova, Italy 
52 	(a)(b)
E. Andronikashvili Institute of Physics, Iv. Javakhishvili Tbilisi State University, Tbilisi; 
High Energy Physics Institute, Tbilisi State University, Tbilisi, Georgia 

JHEP06(2016)093 

53 	II Physikalisches Institut, Justus-Liebig-Universit¨at Giessen, Giessen, Germany 
54 	SUPA -School of Physics and Astronomy, University of Glasgow, Glasgow, U.K. 
55 	II Physikalisches Institut, Georg-August-Universit¨at, G¨ottingen, Germany 
56 	Laboratoire de Physique Subatomique et de Cosmologie, Universit´e Grenoble-Alpes, CNRS/IN2P3, Grenoble, France 
57 	Department of Physics, Hampton University, Hampton VA, U.S.A. 
58 	Laboratory for Particle Physics and Cosmology, Harvard University, Cambridge MA, U.S.A. 
59 	(a) 
Kirchhoff-Institut f¨ur Physik, Ruprecht-Karls-Universit¨at Heidelberg, Heidelberg; (b) Physikalisches Institut, Ruprecht-Karls-Universit¨at Heidelberg, Heidelberg; (c) ZITI Institut f¨ur technische Informatik, Ruprecht-Karls-Universit¨at Heidelberg, Mannheim, Germany 
60 	Faculty of Applied Information Science, Hiroshima Institute of Technology, Hiroshima, Japan 
61 	(a) 
Department of Physics, The Chinese University of Hong Kong, Shatin, N.T., Hong Kong; (b) Department of Physics, The University of Hong Kong, Hong Kong; (c) Department of Physics, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, China 
62 	Department of Physics, Indiana University, Bloomington IN, U.S.A. 
63 	Institut f¨ur Astro-und Teilchenphysik, Leopold-Franzens-Universit¨at, Innsbruck, Austria 
64 	University of Iowa, Iowa City IA, U.S.A. 
65 	Department of Physics and Astronomy, Iowa State University, Ames IA, U.S.A. 
66 	Joint Institute for Nuclear Research, JINR Dubna, Dubna, Russia 
67 	KEK, High Energy Accelerator Research Organization, Tsukuba, Japan 
68 	Graduate School of Science, Kobe University, Kobe, Japan 
69 	Faculty of Science, Kyoto University, Kyoto, Japan 
70 	Kyoto University of Education, Kyoto, Japan 
71 	Department of Physics, Kyushu University, Fukuoka, Japan 
72 	Instituto de F´ısica La Plata, Universidad Nacional de La Plata and CONICET, La Plata, Argentina 
73 	Physics Department, Lancaster University, Lancaster, U.K. 
74 	(a) 
INFN Sezione di Lecce; (b) Dipartimento di Matematica e Fisica, Universit`a del Salento, Lecce, 
Italy 
75 	Oliver Lodge Laboratory, University of Liverpool, Liverpool, U.K. 
76 	Department of Physics, Joˇzef Stefan Institute and University of Ljubljana, Ljubljana, Slovenia 
77 	School of Physics and Astronomy, Queen Mary University of London, London, U.K. 
78 	Department of Physics,RoyalHollowayUniversity ofLondon, Surrey,U.K. 
79 	Department of Physics and Astronomy, University College London, London, U.K. 
80 	Louisiana Tech University, Ruston LA, U.S.A. 
81 	Laboratoire de Physique Nucl´eaire et de Hautes Energies, UPMC and Universit´e Paris-Diderot and CNRS/IN2P3, Paris, France 
82 	Fysiska institutionen, Lunds universitet, Lund, Sweden 
83 	Departamento de Fisica Teorica C-15, Universidad Autonoma de Madrid, Madrid, Spain 
84 	Institut f¨ur Physik, Universit¨at Mainz, Mainz, Germany 
85 	School of Physics and Astronomy, University of Manchester, Manchester, U.K. 
86 	CPPM, Aix-Marseille Universit´e and CNRS/IN2P3, Marseille, France 
87 	Department of Physics, University of Massachusetts, Amherst MA, U.S.A. 
88 	Department of Physics, McGill University, Montreal QC, Canada 
89 	School of Physics, University of Melbourne, Victoria, Australia 
90 	Department of Physics, The University of Michigan, Ann Arbor MI, U.S.A. 
91 	Department of Physics and Astronomy, Michigan State University, East Lansing MI, U.S.A. 
92 	(a) 
INFN Sezione di Milano; (b) Dipartimento di Fisica, Universit`a di Milano, Milano, Italy 93 B.I. Stepanov Institute of Physics, National Academy of Sciences of Belarus, Minsk, Republic of Belarus 94 National Scientific and Educational Centre for Particle and High Energy Physics, Minsk, Republic of Belarus 95 Department of Physics, Massachusetts Institute of Technology, Cambridge MA, U.S.A. 
JHEP06(2016)093 

96 Group of Particle Physics, University of Montreal, Montreal QC, Canada 
97 P.N. Lebedev Physical Institute of the Russian Academy of Sciences, Moscow, Russia 
98 Institute for Theoretical and Experimental Physics (ITEP), Moscow, Russia 
99 National Research Nuclear University MEPhI, Moscow, Russia 

100 	D.V. Skobeltsyn Institute of Nuclear Physics, M.V. Lomonosov Moscow State University, Moscow, Russia 
101 	Fakult¨at f¨ur Physik, Ludwig-Maximilians-Universit¨at M¨unchen, M¨unchen, Germany 
102 	Max-Planck-Institut f¨ur Physik (Werner-Heisenberg-Institut), M¨unchen, Germany 
103 	Nagasaki Institute of Applied Science, Nagasaki, Japan 
104 	Graduate School of Science and Kobayashi-Maskawa Institute, Nagoya University, Nagoya, Japan 
105 	(a) 
INFN Sezione di Napoli; (b) Dipartimento di Fisica, Universit`a di Napoli, Napoli, Italy 
106 	Department of Physics and Astronomy, University of New Mexico, Albuquerque NM, U.S.A. 
107 	Institute for Mathematics, Astrophysics and Particle Physics, Radboud University Nijmegen/Nikhef, Nijmegen, Netherlands 
108 	Nikhef National Institute for Subatomic Physics and University of Amsterdam, Amsterdam, Netherlands 
109 	Department of Physics, Northern Illinois University, DeKalb IL, U.S.A. 
110 	Budker Institute of Nuclear Physics, SB RAS, Novosibirsk, Russia 
111 	Department of Physics, New York University, New York NY, U.S.A. 
112 	Ohio State University, Columbus OH, U.S.A. 
113 	Faculty of Science, Okayama University, Okayama, Japan 
114 	Homer L. Dodge Department of Physics and Astronomy, University of Oklahoma, Norman OK, 
U.S.A. 115 Department of Physics, Oklahoma State University, Stillwater OK, U.S.A. 116 Palack´y University, RCPTM, Olomouc, Czech Republic 117 Center for High Energy Physics, University of Oregon, Eugene OR, U.S.A. 118 LAL, Univ. Paris-Sud, CNRS/IN2P3, Universit´e Paris-Saclay, Orsay, France 119 Graduate School of Science, Osaka University, Osaka, Japan 120 Department of Physics, University of Oslo, Oslo, Norway 121 Department of Physics, Oxford University, Oxford, U.K. 
122 	(a) 
INFN Sezione di Pavia; (b) Dipartimento di Fisica, Universit`a di Pavia, Pavia, Italy 
123 	Department of Physics, University of Pennsylvania, Philadelphia PA, U.S.A. 
124 	National Research Centre ”Kurchatov Institute” B.P.Konstantinov Petersburg Nuclear Physics Institute, St. Petersburg, Russia 
125 	(a) INFN Sezione di Pisa; (b) Dipartimento di Fisica E. Fermi, Universit`a di Pisa, Pisa, Italy 
126 	Department of Physics and Astronomy, University of Pittsburgh, Pittsburgh PA, U.S.A. 
127 	(a) Laborat´orio de Instrumenta¸c˜ao e F´ısica Experimental de Part´ıculas -LIP, Lisboa; (b) Faculdade de Ciˆencias, Universidade de Lisboa, Lisboa; (c) Department of Physics, University of Coimbra, Coimbra; (d) Centro de F´ısica Nuclear da Universidade de Lisboa, Lisboa; (e) Departamento de Fisica, Universidade do Minho, Braga; (f) Departamento de Fisica Teorica y del Cosmos and CAFPE, Universidad de Granada, Granada (Spain); (g) Dep Fisica and CEFITEC of Faculdade de Ciencias e Tecnologia, Universidade Nova de Lisboa, Caparica, Portugal 
128 	Institute of Physics, Academy of Sciences of the Czech Republic, Praha, Czech Republic 
129 	Czech Technical University in Prague, Praha, Czech Republic 
130 	Faculty of Mathematics and Physics, Charles University in Prague, Praha, Czech Republic 
131 	State Research Center Institute for High Energy Physics (Protvino), NRC KI, Russia 
132 	Particle Physics Department, Rutherford Appleton Laboratory, Didcot, U.K. 
133 	(a) INFN Sezione di Roma; (b) Dipartimento di Fisica, Sapienza Universit`a di Roma, Roma, Italy 
134 	(a) 
INFN Sezione di Roma Tor Vergata; (b) Dipartimento di Fisica, Universit`a di Roma Tor Vergata, Roma, Italy 
135 	(a) 
INFN Sezione di Roma Tre; (b) Dipartimento di Matematica e Fisica, Universit`a Roma Tre, Roma, Italy 
JHEP06(2016)093 

136 	(a) 
Facult´e des Sciences Ain Chock, R´eseau Universitaire de Physique des Hautes Energies Universit´e Hassan II, Casablanca; (b) Centre National de l’Energie des Sciences Techniques Nucleaires, Rabat; (c) Facult´e des Sciences Semlalia, Universit´e Cadi Ayyad, LPHEA-Marrakech; 
(d) Facult´e des Sciences, Universit´e Mohamed Premier and LPTPM, Oujda; (e) Facult´e des sciences, Universit´e Mohammed V, Rabat, Morocco 
137 	DSM/IRFU (Institut de Recherches sur les Lois Fondamentales de l’Univers), CEA Saclay (Commissariat `
a l’Energie Atomique et aux Energies Alternatives), Gif-sur-Yvette, France 138 Santa Cruz Institute for Particle Physics, University of California Santa Cruz, Santa Cruz CA, 
U.S.A. 139 Department of Physics, University of Washington, Seattle WA, U.S.A. 140 Department of Physics andAstronomy,University of Sheffield, Sheffield,U.K. 141 Department of Physics, Shinshu University, Nagano, Japan 142 Fachbereich Physik, Universit¨at Siegen, Siegen, Germany 143 Department of Physics, Simon Fraser University, Burnaby BC, Canada 144 SLAC National Accelerator Laboratory, Stanford CA, U.S.A. 
145 	(a) 
Faculty of Mathematics, Physics & Informatics, Comenius University, Bratislava; (b) Department of Subnuclear Physics, Institute of Experimental Physics of the Slovak Academy of Sciences, Kosice, Slovak Republic 
146 	(a) 
Department of Physics, University of Cape Town, Cape Town; (b) Department of Physics, University of Johannesburg, Johannesburg; (c) School of Physics, University of the Witwatersrand, Johannesburg, South Africa 
147 	(a) 
Department of Physics, Stockholm University; (b) The Oskar Klein Centre, Stockholm, Sweden 
148 	Physics Department, Royal Institute of Technology, Stockholm, Sweden 
149 	Departments of Physics & Astronomy and Chemistry, Stony Brook University, Stony Brook NY, 
U.S.A. 150 Department of Physics and Astronomy, University of Sussex, Brighton, U.K. 151 School of Physics, University of Sydney, Sydney, Australia 152 Institute of Physics, Academia Sinica, Taipei, Taiwan 153 Department of Physics, Technion: Israel Institute of Technology, Haifa, Israel 154 Raymond and Beverly Sackler School of Physics and Astronomy, Tel Aviv University, Tel Aviv, 
Israel 
155 	Department of Physics, Aristotle University of Thessaloniki, Thessaloniki, Greece 
156 	International Center for Elementary Particle Physics and Department of Physics, The University 
of Tokyo, Tokyo, Japan 157 Graduate School of Science and Technology, Tokyo Metropolitan University, Tokyo, Japan 158 Department of Physics, Tokyo Institute of Technology, Tokyo, Japan 159 Department of Physics, University of Toronto, Toronto ON, Canada 
160 	(a) 
TRIUMF, Vancouver BC; (b) Department of Physics and Astronomy, York University, Toronto ON, Canada 161 Faculty of Pure and Applied Sciences, and Center for Integrated Research in Fundamental Science 
and Engineering, University of Tsukuba, Tsukuba, Japan 
162 	Department of Physics and Astronomy, Tufts University, Medford MA, U.S.A. 
163 	Department of Physics and Astronomy, University of California Irvine, Irvine CA, U.S.A. 
164 	(a) 
INFN Gruppo Collegato di Udine, Sezione di Trieste, Udine; (b) ICTP, Trieste; (c) 
Dipartimento di Chimica, Fisica e Ambiente, Universit`a di Udine, Udine, Italy 
165 	Department of Physics and Astronomy, University of Uppsala, Uppsala, Sweden 
166 	Department of Physics, University of Illinois, Urbana IL, U.S.A. 
167 	Instituto de F´ısica Corpuscular (IFIC) and Departamento de F´ısica At´omica, Molecular y Nuclear and Departamento de Ingenier´ıa Electr´onica and Instituto de Microelectr´onica de Barcelona (IMB-CNM), University of Valencia and CSIC, Valencia, Spain 
168 	Department of Physics, University of British Columbia, Vancouver BC, Canada 
169 	Department of Physics and Astronomy, University of Victoria, Victoria BC, Canada 
JHEP06(2016)093 

170 Department of Physics, University of Warwick, Coventry, U.K. 171 Waseda University, Tokyo, Japan 172 Department of Particle Physics, The Weizmann Institute of Science, Rehovot, Israel 173 Department of Physics, University of Wisconsin, Madison WI, U.S.A. 174 Fakult¨at f¨ur Physik und Astronomie, Julius-Maximilians-Universit¨at, W¨urzburg, Germany 175 Fakult¨at f¨ur Mathematik und Naturwissenschaften, Fachgruppe Physik, Bergische Universit¨at 
Wuppertal, Wuppertal, Germany 176 Department of Physics, Yale University, New Haven CT, U.S.A. 177 Yerevan Physics Institute, Yerevan, Armenia 178 Centre de Calcul de l’Institut National de Physique Nucl´eaire et de Physique des Particules 
(IN2P3), Villeurbanne, France 
a Also at Department of Physics, King’s College London, London, U.K. 
b Also at Institute of Physics, Azerbaijan Academy of Sciences, Baku, Azerbaijan 
c Also at Novosibirsk State University, Novosibirsk, Russia 
d Also at TRIUMF, Vancouver BC, Canada 
e Also at Department of Physics, California State University, Fresno CA, U.S.A. 
f Also at Department of Physics, University of Fribourg, Fribourg, Switzerland 
g Also at Departament de Fisica de la Universitat Autonoma de Barcelona, Barcelona, Spain 
h Also at Departamento de Fisica e Astronomia, Faculdade de Ciencias, Universidade do Porto, 

Portugal 
i Also at Tomsk State University, Tomsk, Russia 
j Also at CPPM, Aix-Marseille Universit´e and CNRS/IN2P3, Marseille, France 
k Also at Universita di Napoli Parthenope, Napoli, Italy 
l Also at Institute of Particle Physics (IPP), Canada 

m Also at Particle Physics Department, Rutherford Appleton Laboratory, Didcot, U.K. 
n Also at Department of Physics, St. Petersburg State Polytechnical University, St. Petersburg, 
Russia 

o Also at Louisiana Tech University, Ruston LA, U.S.A. 
p Also at Institucio Catalana de Recerca i Estudis Avancats, ICREA, Barcelona, Spain 
q Also at Department of Physics, The University of Michigan, Ann Arbor MI, U.S.A. 
r Also at Graduate School of Science, Osaka University, Osaka, Japan 
s Also at Department of Physics, National Tsing Hua University, Taiwan 
t Also at Department of Physics, The University of Texas at Austin, Austin TX, U.S.A. 
u Also at Institute of Theoretical Physics, Ilia State University, Tbilisi, Georgia 
v Also at CERN, Geneva, Switzerland 

w Also at Georgian Technical University (GTU),Tbilisi, Georgia 
x Also at Ochadai Academic Production, Ochanomizu University, Tokyo, Japan 
y Also at Manhattan College, New York NY, U.S.A. 
z Also atHellenicOpenUniversity, Patras,Greece 

aa Also at Institute of Physics, Academia Sinica, Taipei, Taiwan ab Also at LAL, Univ. Paris-Sud, CNRS/IN2P3, Universit´e Paris-Saclay, Orsay, France ac Also at Academia Sinica Grid Computing, Institute of Physics, Academia Sinica, Taipei, Taiwan ad Also at School of Physics, Shandong University, Shandong, China ae Also at Moscow Institute of Physics and Technology State University, Dolgoprudny, Russia af Also at Section de Physique, Universit´e de Gen`eve, Geneva, Switzerland ag Also at International School for Advanced Studies (SISSA), Trieste, Italy ah Also at Department of Physics and Astronomy, University of South Carolina, Columbia SC, U.S.A. ai Also at School of Physics and Engineering, Sun Yat-sen University, Guangzhou, China aj Also at Institute for Nuclear Research and Nuclear Energy (INRNE) of the Bulgarian Academy of 
Sciences, Sofia, Bulgaria 
ak Also at Faculty of Physics, M.V.Lomonosov Moscow State University, Moscow, Russia 

JHEP06(2016)093 

al Associated at Durham University, IPPP, Durham, U.K. 
am Also at National Research Nuclear University MEPhI, Moscow, Russia an Also at Department of Physics, Stanford University, Stanford CA, U.S.A. ao Also at Institute for Particle and Nuclear Physics, Wigner Research Centre for Physics, Budapest, Hungary ap Also at Flensburg University of Applied Sciences, Flensburg, Germany aq Also at University of Malaya, Department of Physics, Kuala Lumpur, Malaysia 
∗ Deceased 
― 82 ― 

JHEP06(2016)093