View
5
Download
0
Category
Preview:
Citation preview
Jesse Thaler — Jet Studies for Les Houches 1
Jesse Thaler
Les Houches Workshop — June 6, 2017
Jet Studies for
Les Houches
Jesse Thaler — Jet Studies for Les Houches 2
Question to Les Houches royalty: !
First dedicated jet session at LH 2007 (Salam/Wobisch)?First official “jet convener” at LH 2011 (Soyez)?
[JADE,1979][Hanson, et al., 1975]
Over Four Decades of Jets and QCD!
Jesse Thaler — Jet Studies for Les Houches 3
100 MeV++
1.3 GeV++
4.2 GeV++
0++
u,d,s
c
b
g
•
•Jets from theStandard Model
micro-jet≈ 65%
++ = plus gluonic radiation
Jesse Thaler — Jet Studies for Les Houches 4
100 MeV++
1.3 GeV++
4.2 GeV++
0++
173 GeV++
80/91 GeV
u,d,s
c
b
g
•
•
t•
W/Z
H ••
125 GeV
Jets from theStandard Model
micro-jet≈ 70%
≈ 70%
≈ 60%
≈ 65%
++ = plus gluonic radiation
Jesse Thaler — Jet Studies for Les Houches 5
100 MeV++
1.3 GeV++
4.2 GeV++
0++
173 GeV++
80/91 GeV
u,d,s
c
b
g
•
•
t•
W/Z
H ••
125 GeV
≈ 70%
≈ 70%
≈ 60%Jet Grooming
⇒
Jet Discrimination
vs.
Substructure Toolbox:
Jesse Thaler — Jet Studies for Les Houches 6
0.4 0.3 0.2 0.1 0 0.1 0.2 0.3 0.4
Events
1−10
1
10
210
310
410
510
610
710QCD W+jets
+jetstt Z+jets
Z'(50) Z'(100)
Z'(150) Z'(200)
Data
(13 TeV)-1
35.9 fbCMS Preliminary
1,DDT
2 AK8 N
0.4− 0.3− 0.2− 0.1− 0 0.1 0.2 0.3 0.4
Data
/Sim
ula
tion
0
1
2
Nu
mb
er
of
eve
nts
/5 G
eV
0
200
400
600
800
1000
1200
1400
1600
1800
2000
2200
2400Data
Multijet pred.
Total SM pred.
qq+jets→W
qq+jets→Z
=135 GeVZ'
=1/6, mq
Z'(qq), g
(13 TeV)-1
35.9 fbCMS Preliminary
: 700-800 GeVT
p
(GeV)PUPPI
SDAK8 m
60 80 100 120 140 160 180 200 220 240
Data
/Pre
dic
tion
0.9
1
1.1
E.g. CMS Boosted Z’ Search
Recoil Jet(s)
with N2 + Decorrelation + Soft Drop + PUPPI
pp
Z or Z’ ?
[CMS PAS EXO-17-001;using Moult, Necib, JDT, 1609.07483; Dolen, Harris, Marzani, Rappoccio, Tran, 1603.00027;Larkoski, Marzani, Soyez, JDT, 1402.2657; Bertolini, Harris, Low, Tran, 1407.6013]
Z’-like q/g-like
Jesse Thaler — Jet Studies for Les Houches 7
Back to the Future
Z' mass (GeV)50 60 100 200 300 400 1000
qcouplin
g, g
0.02
0.03
0.04
0.1
0.2
0.3
0.4
1CMSPreliminary (13 TeV)
-135.9 fb
UA2
CDF Run 1
CDF Run 2
Observed
Expected
1 std. deviation±
2 std. deviation±
γ, ISR -1ATLAS13, 15.5 fb
, ISR jet-1ATLAS13, 15.5 fb
, Scouting-1CMS8, 18.8 fb
Z Width (indirect)
[CMS PAS EXO-17-001]
Jet Physics = Innovative Approaches to Hadronic Final States
1993
2017
Yes, searching for abump on a bump!
Jesse Thaler — Jet Studies for Les Houches 8
Back to Les Houches 2015
Jesse Thaler — Jet Studies for Les Houches 9
Jesse Thaler — Report of the Les Houches Quark/Gluon Subgroup
Jesse Thaler
Les Houches Workshop — June 10, 2015
Report of the Les Houches Quark/Gluon Subgroup
1
on behalf of Andy Buckley, Jon Butterworth, Mario Campanelli, Marat Freytsis, Peter Loch, Philippe Gras, Deepak Kar, Simon Plätzer, Andrzej Siodmok,
Peter Skands, Dave Soper, Gregory Soyez, Frank Tackmann
= Jets are Les H.
Jesse Thaler — Jet Studies for Les Houches 10
Jesse Thaler — Report of the Les Houches Quark/Gluon Subgroup 2
[Larkoski, JDT, Waalewijn,1408.3122] [based on Berger, Kucs, Sterman, hep-ph/0303051; Ellis, Vermilion, Walsh, Hornig, Lee, 1001.0014]
[see also Larkoski, Salam, JDT, 1305.0007; Larkoski, Neill, JDT, 1401.2158] [For a more complete catalog, see Gallicchio, Schwartz, 1106.3076, 1211.7038]
0 1 2
λκβ
κ
β0
1
2 pD
T
eβ
width mass
multiplicity
λκβ =
∑
i∈jet
zκiθβi
momentum fraction
angle to recoil-free axis
Hunting the White Whale: Quarks vs. Gluons
Probe radiation pattern with e.g. Generalized Angularities
Quark: CF = 4/3 Gluon: CA = 3vs.
Cartoon:
zi
θi
LHA (Les Houches Angularity)
Jesse Thaler — Jet Studies for Les Houches 11
Coming soon to JHEP!
Jesse Thaler — Jet Studies for Les Houches 12
Idealized Quark/Gluon Distributions
Large variations for gluon jets (not tuned to LEP)
0
1
2
3
4
5
6
0 0.2 0.4 0.6 0.8 1
Q=200 GeV
R=0.6
pg(1 0.5)
1 [LHA]
Gluon, hadron-level
Pythia 8.215Herwig 2.7.1Sherpa 2.2.1Vincia 2.001
Deductor 1.0.2Ariadne 5.0.
Dire 1.0.0Analytic NLL
0
1
2
3
4
5
6
0 0.2 0.4 0.6 0.8 1
Q=200 GeV
R=0.6
pq(1 0.5)
1 [LHA]
Pythia 8.215Herwig 2.7.1Sherpa 2.2.1Vincia 2.001
Deductor 1.0.2Ariadne 5.0.
Dire 1.0.0Analytic NLL
e+e– → quarks (CF = 4/3) e+e– → gluons (CA = 3)
LHA =
X
i
zi
p
θi
vs.
[Gras, Hoeche, Kar, Larkoski, Lönnblad, Plätzer, Siódmok, Skands, Soyez, JDT, 1704.03878;based on Soyez, JDT, Freytsis, Gras, Kar, Lönnblad, Plätzer, Siodmok, Skands, Soper, 1605.04692]
LHA =
X
i
zi
p
θi
Q = 200 GeVR = 0.6
Q = 200 GeVR = 0.6
Jesse Thaler — Jet Studies for Les Houches 13
Impact on Quark/Gluon Separation
0
0.5
1
1.5
2
0 0.2 0.4 0.6 0.8 1
Q=200 GeV
R=0.6d/d1 0.5
1 [LHA]
Separation, hadron-level
Pythia 8.215Herwig 2.7.1Sherpa 2.2.1Vincia 2.001
Deductor 1.0.2Ariadne 5.0.
Dire 1.0.0Analytic NLL
Differential Integrated Values
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
(0,0) (2,0) (1,0.5) (1,1) (1,2)
Q=200 GeV
R=0.6
LL
Separation:
Angularity: ( , )
Hadron-level
Pythia 8.215Herwig 2.7.1Sherpa 2.2.1Vincia 2.001
Deductor 1.0.2Ariadne 5.0.
Dire 1.0.0Analytic NLL
Affects both IRC unsafe and IRC safe observables
LHA =
X
i
zi
p
θi
Q = 200 GeVR = 0.6
λ1
2
(
Q(λ)−G(λ))2
Q(λ) +G(λ)
Jesse Thaler — Jet Studies for Les Houches 14
Testing Shower Variants
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
(0,0) (2,0) (1,0.5) (1,1) (1,2)
Q=200 GeV
R=0.6
LL
Separation:
Angularity: ( , )
Ariadne 5.0. , hadron-level
baselineno g qqno swing
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
(0,0) (2,0) (1,0.5) (1,1) (1,2)
Q=200 GeV
R=0.6
LL
Separation:
Angularity: ( , )
Dire 1.0.0, hadron-level
baselineno g qqMC@NLO1-loop s
3-loop s
string had.
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
(0,0) (2,0) (1,0.5) (1,1) (1,2)
Q=200 GeV
R=0.6
LL
Separation:
Angularity: ( , )
Herwig 2.7.1, hadron-level
baselineno g qqno CR
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
(0,0) (2,0) (1,0.5) (1,1) (1,2)
Q=200 GeV
R=0.6
LL
Separation:
Angularity: ( , )
Pythia 8.215, hadron-level
baselineno g qqno ME
2-loop s
CR1
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
(0,0) (2,0) (1,0.5) (1,1) (1,2)
Q=200 GeV
R=0.6
LL
Separation:
Angularity: ( , )
Sherpa 2.2.1, hadron-level
baselineno g qqNjet=1Njet=0
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
(0,0) (2,0) (1,0.5) (1,1) (1,2)
Q=200 GeV
R=0.6
LL
Separation:
Angularity: ( , )
Vincia 2.001, hadron-level
baselineno g qqno ME
1-loop s
alt QE
Large impact from color reconnection
Σ(eβ) =e−γER0(eβ)
Γ (1 +R0(eβ))e−R(eβ) e
−fNGL(eβ)
Jesse Thaler — Jet Studies for Les Houches 15
NLL Resummation plus NP Correction
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
(0,0) (2,0) (1,0.5) (1,1) (1,2)
Q=200 GeV
R=0.6
LL
Separation:
Angularity: ( , )
Analytic NLL, hadron-level
baselineno g qqno NGLs1-loop s
no Ci in 0
no Ci in 0
alt F( ) F () =
220exp
−
4
2
20
]
) = 0.
Z
0
Z
0
=1
β − 1
Ω0
REJ
1−
Ξ0
REJ
β−1!
Ωg0
Ωq0
=Ξg0
Ξq0
=CA
CF.
NLL with Non-Global Logs
Shift from NP Shape Function
Convolution (Simplified)
Default: Full Casimir Scaling↑: No Casimir Scaling for Ξ0 ↓: No Casimir Scaling for Ω0 or Ξ0
dσ
deβ=
Z
deβ d σ(eβ)F () δ (eβ − eβ − )
Large impact from nonperturbative assumptions
Jesse Thaler — Jet Studies for Les Houches 16
Differences Persist at LHC
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0 0.2 0.4 0.6 0.8 1 1.2
pTmin=100 GeV
Separation:
R
0.5 [LHA], Hadron-level, plain jet
Pythia 8.215Herwig 2.7.1Sherpa 2.2.1Vincia 2.001
LHC sensitive to jet radiationpatterns (esp. gluons)
0
1
2
3
4
5
0 0.2 0.4 0.6 0.8 1
pTmin=100 GeV
R=0.4p2j(1 0.5)
10.5 [LHA]
2j, Hadron-level, plain jet
Pythia 8.215Herwig 2.7.1Sherpa 2.2.1Vincia 2.001
0
1
2
3
4
5
0 0.2 0.4 0.6 0.8 1
pTmin=100 GeV
R=0.4pZj(1 0.5)
10.5 [LHA]
Z+j, Hadron-level, plain jet
Pythia 8.215Herwig 2.7.1Sherpa 2.2.1Vincia 2.001
pp → Z + jet85% quarks @ 100 GeV
pp → dijet75% gluons @ 100 GeV
Varying Jet Radius
Jesse Thaler — Jet Studies for Les Houches 17
Recommendations
[Gras, Hoeche, Kar, Larkoski, Lönnblad, Plätzer, Siódmok, Skands, Soyez, JDT, 1704.03878;based on Soyez, JDT, Freytsis, Gras, Kar, Lönnblad, Plätzer, Siodmok, Skands, Soper, 1605.04692]
Scrutinize aspects of final-state parton shower!
Gluon splitting to quark/antiquarkColor reconnection models
Default choice of αs
!
New tuning/measurement campaigns !
LEP: Tune to existing e+e– → b b g measurementsLEP: Measure Les Houches Angularity (LHA, β = 1/2)
LHC: Differential studies in multiple channels at multiple pT and R!
Further analytic studies !
Pursue (N)NLL + (N)NLO calculationsFirst-principles understanding of NP corrections?
_
Jesse Thaler — Jet Studies for Les Houches 18
Assess matching/merging uncertainties and impact on jet substructure methods (and vice versa)
Action Items for LH 2017Overlap with other working groups
Develop concrete plan for LHC (and LEP) jet shapemeasurements and parton shower tuning (see backup)
Identify new uses for quark/gluon tagging,(e.g. improved PDF extraction?)
Beyond quark/gluon discrimination, correct modeling ofjet radiation will benefit many LHC studies
Jesse Thaler — Jet Studies for Les Houches 19
Further Goals for Les Houches 2017
Jesse Thaler — Jet Studies for Les Houches 20
Precision Jet Substructure Calculations!
Improved control with jet grooming (e.g. top mass, αs extraction)Quantifying impact of soft QCD effects (underlying event and hadronization)
Charmonium polarization puzzle in identified jetsHeavy ion physics: jet shapes in medium, probe QGP with boosted objects
…!
Advanced Tagging Strategies!
Interplay of jet radius, jet discriminants, and jet grooming (e.g. dichroic)Exploiting charm/bottom tags (e.g. gluon splitting to heavy flavor)
More refined tagging categories (e.g. longitudinal vs. transverse W/Z bosons)Machine learning: multi-category classification, physics-inspired network design
…!
Many points of contact with other working groups
Two Broad ThemesFrom the wiki and pre-LH discussions
Jesse Thaler — Jet Studies for Les Houches 21
Deep Thinking meets Deep Learning
Starting from well-defined physics goal…
Blackboards, Coffee,Mountains, Inspiration
Validation inSimulation/Data/Calculations
⇒
Construct ExplicitOptimization Problem
Training inSimulation/Data
⇒⇔
Jesse Thaler — Jet Studies for Les Houches 22
Deep Thinking meets Deep Learning
0.0 0.1 0.2 0.3 0.40
2
4
6
8
10
12
N2(2)
ProbabilityDensity
N2(2)
Pythia 8.219
R=1.0, pT>500 GeV, m [80, 100] GeV
ZLight QuarkGluon
[e.g. Moult, Necib, JDT, 1609.07483; Datta, Larkoski, 1704.08249; de Oliveira, Kagan, Mackey, Nachman, Schwartzman, 1511.05190]
Energy Correlatorsfrom Power Counting
ConvolutionalNeural Network
M-Body Discrimination
13 TeV, pT > 500 GeV, R = 0.8
Pythia8
+ -
+ - + -
+ - + -
Dense Network ofN-subjettiness Values
Continuum of approaches, ripe for cross fertilization
Optimize for performance and robustness (and calculability)
vs.q/g W/Ze.g.
Jesse Thaler — Jet Studies for Les Houches 23
My goal for Les Houches:Hunting the next white whale for jet physics
Jesse Thaler — Jet Studies for Les Houches 24
Extracting Strong Coupling Constant
Determine αs from
LHC jet substructure?
Clarify role of jet groomingfor precision studies?!Highlight ways to controlsoft QCD corrections?!Inform strategies usedat lepton colliders?!Ignite progress in(N)NLO substructure? Needed to get in the PDG!
Jesse Thaler — Jet Studies for Les Houches 25
Proof of ConceptSoft-dropped jet mass
⇒ mg
Grooming simplifies structure of calculation, reduces NP effects!
[Frye, Larkoski, Schwartz, Yan, 1603.06375, 1603.09338; see also Marzani, Schunk, Soyez, 1704.02210]
First NNLL + O(αs2) result for substructure in pp (!)
/
- -
( )
= =
+ > =
++ ( + )
++ ( + )
Jesse Thaler — Jet Studies for Les Houches 26
0
0.05
0.1
0.15
0.2
0.25
10-6
10-5
10-4
10-3
10-2
10-1
100
C1(2)
R=1, pt>3 TeVzcut=0.1
plain jetβ=2β=1β=0β=-0.5
Pythia 8
mg2/pT
2
β < 0 β = 0 β > 0
Less Grooming
β → ∞β → –∞
More Grooming
![Larkoski, Marzani, Soyez, JDT, 1402.2657; see also Butterworth, Davison, Rubin, Salam, 0802.2470; Dasgupta, Fregoso, Marzani, Salam, 1307.0007]
0
0.05
0.1
0.15
0.2
0.25
10-6
10-5
10-4
10-3
10-2
10-1
100
C1(2)
R=1, pt>3 TeVzcut=0.1
solid: mult. em.plain jetβ=2β=1β=0β=-0.5
NLL Resummation
mg2/pT
2
Sudakov factor
Kink from zcut
⇒ mg
Systematic Strategy for αs?Scanning soft-drop variants?
Jesse Thaler — Jet Studies for Les Houches 27
for jet vetos
N-subjettiness forboosted objects
N-jettiness forNNLO calculations
[Stewart, Tackmann, Waalewijn, 1004.2489; Kim, 1011.1493; JDT, Van Tilburg, 1011.2268,1108.2701; Boughezal, Focke, Liu, Petriello, 1504.02131;Gaunt, Stahlhofen, Tackmann, Walsh, 1505.04794; Stewart, Tackmann, JDT, Vermilion, Wilkason, 1508.01516; JDT, Wilkason, 1508.01518]
Jet 2
Jet b Jet a
Soft
Jet 3
Jet 1b
a
1
32
p p
ℓ−
ℓ+
0 0.2 0.4 0.6 0.8 10
0.01
0.02
0.03
0.04
0.05
0.06
3/
2 of jet
Rela
tive o
ccure
nce
145 GeV < mj < 205 GeV
Top jets
QCD jets
N-jettiness
XCone for exclusive cone jets
Benefits of Cross-pollinationwide-angle ⇔ boosted regimes
Jesse Thaler — Jet Studies for Les Houches 28
Jet Studies for Les Houches
Pursuing white whale of quark/gluon discrimination
reveals (non)perturbative uncertainties in jet radiation
2015:
Pursuing _______________________________
!
reveals __________________________________
2017:
Looking forward to a fun, productive workshop!
(topic of interest to jet physics community)
(insights with broader implications for QCD and beyond)
Jesse Thaler — Jet Studies for Les Houches 29
Backup Slides
Jesse Thaler — Jet Studies for Les Houches 30
Innovative Approaches to Hadronic Final States
SearchesMeasurements
(GeV)T
Leading top p0 200 400 600 800 1000 1200
(pb/G
eV
)(1
)
T/d
pσ
d
4−10
3−10
2−10
1−10
1
10
210
Parton level
Data (resolved)
Data (boosted)
Powheg
aMC@NLO
Madgraph
(13 TeV)-12.53 fb
CMS
Preliminary
[e.g. CMS-PAS-JME-15-002; CMS-PAS-TOP-16-013; ATLAS-CONF-2016-014][e.g. Lapsien, Kogler, Haller, 1606.04961; Larkoski, Moult, 1510.08459; Schwaller, Stolarski, Weiler, 1502.05409]
Z' mass [TeV]0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
) [p
b]
t t
→ B
R(Z
' ×
Z'
σ
2−10
1−10
1
10
210
310
Observed 95% CL limit
Expected 95% CL limit
uncertaintyσExp. 1
uncertaintyσExp. 2
1.3)×/m=1.2%) (LO Γ(TC2
Z'
1.3)×/m=3%) (LO Γ(TC2
Z'
ATLAS Preliminary
-1 = 13 TeV, 3.2 fbs
Sε
0 0.2 0.4 0.6 0.8 1
Bε
4−10
3−10
2−10
1−10
CMSSimulation Preliminary 13 TeV
| < 1.5η < 1000 GeV, |T
800 < p
R(top,parton) < 0.6∆
η and T
flat p
CMSTT min. m
CMSTT top m
Filtered (r=0.2, n=3) m
RecHTT V2 f
HTT V2 m
=0.5) mcut
Pruned (z=0.1, r
Q-jet volatility
=0) mβSoftdrop (z=0.1,
=1) mβSoftdrop (z=0.2,
Trimmed (r=0.2, f=0.03) m
2τ/
3τUngroomed
) (R=0.2)χlog(
φ
3− 2− 1− 0 1 2 3
η
3−
2−
1−
0
1
2
3
HOTVREvent 2clustered with HOTVR
Performance
-
-
-
( )
( )
( )=
/
( )
( )
( )
+ -
[ ]
Strategies Calculations Scenarios
Jesse Thaler — Jet Studies for Les Houches 31
Jesse Thaler — Report of the Les Houches Quark/Gluon Subgroup 3
What is a Quark Jet? From lunch/dinner discussions
A quark parton
A Born-level quark parton
The initiating quark parton in a final state shower
An eikonal line with baryon number 1/3and carrying triplet color charge
A quark operator appearing in a hard matrix elementin the context of a factorization theorem
A parton-level jet object that has been quark-tagged using a soft-safe flavored jet algorithm (automatically collinear safe if you sum constituent flavors)
A phase space region (as defined by an unambiguous hadronic fiducial cross section measurement) that yields an enriched sample of quarks (as interpreted by some suitable, though fundamentally ambiguous, criterion)
Ill-Defined
Well-Defined What we mean
What people
sometimes
think we mean
Quark
as adjective
Quark
as noun
Jesse Thaler — Jet Studies for Les Houches 32
Systematic Differences between Generators
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
50 100 200 500 1000
R=0.6
s= s0
LL
Separation:
Q [GeV]
0.5 [LHA], hadron-level
Pythia 8.215Herwig 2.7.1Sherpa 2.2.1Vincia 2.001
Deductor 1.0.2Ariadne 5.0.
Dire 1.0.0Analytic NLL
0
0.05
0.1
0.15
0.2
0.25
0.3
0.35
0.4
0.45
0 0.2 0.4 0.6 0.8 1 1.2
Q=200 GeV
s= s0
LL
Separation:
R
0.5 [LHA], hadron-level
Pythia 8.215Herwig 2.7.1Sherpa 2.2.1Vincia 2.001
Deductor 1.0.2Ariadne 5.0.
Dire 1.0.0Analytic NLL
Differences in both absolute and relative trends
Varying Collision Energy Varying Jet Radius
Jesse Thaler — Jet Studies for Les Houches 33
Proposed Systematic LHC Study
Strategy:
Observables:!
Samples:!
!
!
Variations:
multiplicity, pTD
, LHA, width, thrust, …
!
V + jet = quark-enrichedvs. dijet = gluon-enriched(No need to diagonalize?)!
Multiple jet pT rangesMultiple jet radii e.g. R = 0.2, 0.4, 0.6, 0.8,1.0With & without grooming e.g. mMDT (soft drop β=0, zcut=0.1)
+
+
+
+
( )
(Over)constrain final-state parton shower
Key task for jet physics community
In principle, benefits every LHC measurement
Recommended