User Tools

Site Tools


2023:topics

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
2023:topics [2023/03/08 17:21]
stephen.jones [Session 1]
2023:topics [2023/06/22 16:06] (current)
gregor.kasieczka [Session 2]
Line 1: Line 1:
 ===== List of topics for 2023 ====== ===== List of topics for 2023 ======
 //​(accessible to everyone, editing rights to conveners and organizers)//​ //​(accessible to everyone, editing rights to conveners and organizers)//​
 +---------------------------
  
 ==== Session 1 ==== ==== Session 1 ====
Line 59: Line 60:
   * Miscellaneous   * Miscellaneous
      * forward physics -> FASER. Anything needed from SM point of view?      * forward physics -> FASER. Anything needed from SM point of view?
 +
 +  * Machine Learning
 +     * Matrix Element calculation using ML
 +     * Interpretable models
 +     * Fast surrogate models for physics simulations ​
 +     * Workflows and interoperability with experimental software ​
 +     * Incorporating uncertainties in the training of ML models
 +     * ML-based unfolding techniques
 +     * Enforcing properties to ML models: Lorentz invariance/​equivariance,​ permutation invariance, IRC safety
 +
  
  
Line 65: Line 76:
   * starting from state-of-the-art ML-taggers, estimate how much they are correlated to the physics in the measurements of substructure we already have. Discuss good procedures to best tune MC to describe ML-taggers. Discuss good procedures to estimate uncertainties on ML-taggers, given the physics understanding of various effects (an example is scale-variations,​ where our constraints from data are certainly stronger than the factor 2 variations in many cases, but sometimes the opposite).   * starting from state-of-the-art ML-taggers, estimate how much they are correlated to the physics in the measurements of substructure we already have. Discuss good procedures to best tune MC to describe ML-taggers. Discuss good procedures to estimate uncertainties on ML-taggers, given the physics understanding of various effects (an example is scale-variations,​ where our constraints from data are certainly stronger than the factor 2 variations in many cases, but sometimes the opposite).
   * take stock of how well most recent generator developments have improved description of multiple new measurements of substructure. (Followup from a previous Les Houches 2015 on q/​g-tagging)   * take stock of how well most recent generator developments have improved description of multiple new measurements of substructure. (Followup from a previous Les Houches 2015 on q/​g-tagging)
 +
 +
 +**__Tools, Event Generators and Machine Learning:​__**
  
  
Line 132: Line 146:
 ==== Session 2 ==== ==== Session 2 ====
  
-....+**__Higgs:​__**
  
 +  * "​Precision":​
 +    * Trilinear and quadrilinear Higgs couplings
 +    * Lepton flavour violation in the Higgs sector
 +    * Constraining the CP structure of the Higgs couplings
 +    * Higgs and EFT
 +    * Characterizing Higgs boson production and decay
 +  * "​Novelties":​
 +    * Exotic decays of the 125 GeV Higgs boson
 +    * Additional Higgs bosons (low/high mass) - uncovered parameter space
 +    * Naturalness
 +       * Gaps in searches for traditional models (SUSY, compositness)
 +       * Searches for cosmological triggers (new Higgs bosons and v-like leptons)
 +       * Unexpected signatures of naturalness
  
----- 
  
-----+**__Machine Learning and Tools:​__** 
 +   * **Reinterpretation**:​ Tools like Rivet, Gambit, .., ML for reinterpretation,​ reinterpretation of model agnostic searches 
 +   * **EFTs**: Fits, PDF inclusion, SMEFT progress ... 
 +   * **Physics & ML**: Injection of physics priors into classification and generative models; interpretable ML techniques; ML for optimal sensitivity for BSM searches and optimal observables and likelihood learning, opening the black box 
 +   * **Anomalies**:​ A unified view of different anomaly detection techniques,​.pushing the boundaries of anomaly detection
  
-==== Session 1 (from 2021) ==== 
-Here for convenience 
  
-**__Jet substructure techniques:​__** 
  
-  ​Interplay of jet substructure with other groups (e.g. jet substructure for EW measurements like VBF, semileptonic VV, etc.) +**__Low-energy precision probes ​of BSM:__** 
-  ​Snowmass jet substructure report (JSS at future colliders) [work + discussion] +   ​New Electric Dipole Moment searches? 
-  * What can we do with previous colliders, in light of the LHC [discussion] +   ​Parity violation ​in new systems? 
-  * Comparing new unfolding methods +   ​Exotic atoms/ions to test exotic forces 
-      * Accord on unbinned results (?) +       ​muonic atoms 
-  * Probing the latest MC generators (PB algorithm, PANScales, Deductor...) with jet substructure +       ​antiprotonic atoms 
-      * Finite N_C, beyond LL, ... +       ​highly charged ions 
- +       ​Rydberg states
- +
- +
-**__Monte Carlo:__** +
-  +
-    ​Non-perturbative uncertainties +
-      common hadronisation interface and variations +
-      * theoretical understanding +
-      * differences ​in tuned comparisons +
-      pheno impact for certain classes of processes (e.g. VBF/VBS) +
-    Shower accuracy studies +
-      comparing different schemes on higher orders, evaluate phenomenological impact +
-      Subleading colour and interplay with colour reconnection +
-      New sampling methods and algorithms versus machine learning techniques +
-      * Accuracy of merging resummed calculation versus ME+PS paradigms +
-    * Photon physics, modelling of fragmentation +
-    * Heavy flavour matching +
-      * review of existing measurements +
-      * Connecting precision calculation,​ fragmentation and decays +
-      * partons at 100 TeV +
-    * Common LHC event bazaar +
-    * Status and needs for electroweak corrections and radiation in shower algorithms +
-    * Machine learning and adaptive Monte Carlo methods +
- +
-==== Session 2 ====+
  
2023/topics.1678292465.txt.gz · Last modified: 2023/03/08 17:21 by stephen.jones