Showing 1–1 of 1 results for author: Vigl, M

Finetuning Foundation Models for Joint Analysis Optimization

Authors: Matthias Vigl, Nicole Hartman, Lukas Heinrich

Abstract: In this work we demonstrate that significant gains in performance and data efficiency can be achieved in High Energy Physics (HEP) by moving beyond the standard paradigm of sequential optimization or reconstruction and analysis components. We conceptually connect HEP reconstruction and analysis to modern machine learning workflows such as pretraining, finetuning, domain adaptation and high-dimensi… ▽ More In this work we demonstrate that significant gains in performance and data efficiency can be achieved in High Energy Physics (HEP) by moving beyond the standard paradigm of sequential optimization or reconstruction and analysis components. We conceptually connect HEP reconstruction and analysis to modern machine learning workflows such as pretraining, finetuning, domain adaptation and high-dimensional embedding spaces and quantify the gains in the example usecase of searches of heavy resonances decaying via an intermediate di-Higgs system to four $b$-jets. △ Less

Submitted 25 January, 2024; v1 submitted 24 January, 2024; originally announced January 2024.

Comments: 13 pages, 12 figures