-
Notifications
You must be signed in to change notification settings - Fork 6
May '22 in Grenoble (SModelS Fest '22)
Sabine Kraml edited this page May 27, 2022
·
10 revisions
Gael Alguero, Mohammad Altakach, Jack Araz, Andy Buckley, Jon Butterworth, Johnathon Gargalionis, Sabine Kraml, Andre Lessa, Timothee Pascal, Humberto Reyes Gonzalez, Ingo Schienbein, Sezen Sekmen, Wolfgang Waltenberger, Jamie Yellen
- Ten years since the Higgs discovery, what's next at the energy frontier?, May 5 by Jon
- Bridge between Classical & Quantum Machine Learning, May 11 by Jack
- Statistically Learning the Next Standard Model from LHC Data, May 19 by Andre and Wolfgang
-
SModelS v2.2.x
- combo_fixes and develop to become official release .... when?
- bug fixes: negative r-values, r_combined >> sum of r's
- do physics study and paper once v2.2.1 is frozen
- new experimental feature: enable trivial covariance matrices
- make experimental features flag a list: llhdsFromLimits, trivialCovMats
-
Learning likelihoods
- main people: HRG, TP, ....
- ML pyhf llhds, provide model for fast llhd evaluation in SModelS
- open issues: portability of e.g. ONXX models; how to store, ship, package, ....
- also need to understand why pyhf llhds are truncated at some mu<0 and whether this should be learned or replaced by extrapolation
-
TACO
- main people: HRG, WW, JY + now JA
- include SK, AL in the effort, perhaps also MA
- test correlations btw searches and measurements
- build simplified per-analysis overlap matrix
- Pathfinder for SModelS and/or protomodel builder
- positive signals mode: anomaly hunt
-
Contur
- main people: MA, JY, AB, AL, ...
- case study of complementary constraints from searches and measurements
- try jets+MET measurements with complex gluino signals, crosscheck with MA5/Checkmate (AL+MA)
- positive signals mode: anomaly hunt (JB et al)
-
Generalized topologies (graphs)
- main people: AL
- should go on top of existing SModelS
- new database format
- map decomposition topologies with unique topologies in the DB
- need 2-3 examples of topologies/results without Z2 assumption
- missing topologies format ???
-
UV completion of protomodels
- main people: JG, AL
- JG to do UV model builder
- provide non-trivial protomodel examples for testing
-
ADL
- main people: SS, WW
- provide standalone validation code for ADL people
- efficiency map production
-
Statistics package
- main people: JA, WW
- task: make standalone repository for likelihood calculations and combinations
- step 1 is refactoring, step 2 integration into SModelS and MA5
- timescale: couple of months