Skip to content
DraconiticResearch
Search Sources API
203,234 studies31 sourcesUpdated 43 min ago
PreprintCC BY 4.0arXiv

Sequential SBI for Resonance Searches at the LHC

  • Henning Bahl
  • Ranit Das
Open at sourcePDFDOICite

Abstract

Neural simulation-based inference has been shown to extract significantly more information from LHC data than histogram-based analyses. So far, it has mostly been applied to couplings, for which the polynomial structure of the scattering amplitude can be exploited. Masses and widths instead move and deform the signal in phase space, and learning their likelihood requires many event samples covering the continuous theory parameter space. We show how sequential simulation-based inference can address this by iteratively concentrating the simulations where the observed data points. Using a parton-level example with an exact reference likelihood and reconstruction-level analyses of A→ ZH→ ℓℓ bb̅ and Y→ HH→ (tt̅)(tt̅), we demonstrate that sequential training produces significantly more accurate and better calibrated likelihoods than single round training with the same amount of training events.

Related

  1. Continuous Event Weighting for Resonance Searches
    Yi Yang, Jun-Jie Chang, Hao-Yin Liao
    arXivPreprintCC BY-NC-ND 4.0Full textComputational physicsHigh-energy & particle physicsarXiv
  2. The era of charmed resonance searches at the LHC
    Daniel Whiteson
    arXivPreprintarXiv licenceFull textHigh-energy & particle physicsarXiv
  3. Follow-up comments to searches for QCD instantons via forward proton tagging
    Marek Tasevsky, Valery A. Khoze, Misha G. Ryskin
    arXivPreprintCC BY 4.0Full textHigh-energy & particle physicsarXiv
  4. Cross Section Measurements and Searches for New Structures in e+e- Annihilations at BESIII
    Hao Liu
    arXivPreprintarXiv licenceFull textHigh-energy & particle physicsarXiv
  5. Tau and dark sector searches at Belle and Belle II
    Wenzhe Li
    arXivPreprintarXiv licenceFull textHigh-energy & particle physicsarXiv

Cite

BibTeXRIS
BibTeX
@misc{bahl2026sequential,
  title = {{Sequential SBI for Resonance Searches at the LHC}},
  author = {Henning Bahl and Ranit Das},
  howpublished = {arXiv},
  year = {2026},
  month = oct,
  doi = {10.48550/arxiv.2610.30245},
  eprint = {2610.30245},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2610.30245},
  note = {CC BY 4.0},
}
RIS
TY  - UNPB
TI  - Sequential SBI for Resonance Searches at the LHC
AU  - Henning Bahl
AU  - Ranit Das
PY  - 2026
DA  - 2026/10/21
PB  - arXiv
DO  - 10.48550/arxiv.2610.30245
UR  - https://arxiv.org/abs/2610.30245
AB  - Neural simulation-based inference has been shown to extract significantly more information from LHC data than histogram-based analyses. So far, it has mostly been applied to couplings, for which the polynomial structure of the scattering amplitude can be exploited. Masses and widths instead move and deform the signal in phase space, and learning their likelihood requires many event samples covering the continuous theory parameter space. We show how sequential simulation-based inference can address this by iteratively concentrating the simulations where the observed data points. Using a parton-level example with an exact reference likelihood and reconstruction-level analyses of A→ ZH→ ℓℓ bb̅ and Y→ HH→ (tt̅)(tt̅), we demonstrate that sequential training produces significantly more accurate and better calibrated likelihoods than single round training with the same amount of training events.
ER  -