Parts | Days | Selection | Search | Updates | Downloads | Help

T: Fachverband Teilchenphysik

T 50: Data, AI, Computing, Electronics V

T 50.4: Talk

Wednesday, March 18, 2026, 17:00–17:15, KH 00.024

Unbinned Unfolding of the WWbb Analysis with OmniFold — •Josef Murnauer, Daniel Britzger, and Stefan Kluth — Max-Planck-Institute for Physics

We revisit the recently published ATLAS measurement of WWbb production and explore an alternative unfolding strategy based on OmniFold. The newly published single- and di-lepton WWbb cross-section measurements have demonstrated that this process constitutes a major new avenue for precision studies in top-quark physics at the LHC. New paradigms in data analysis for Run-III and beyond are emerging rapidly, with traditional techniques (such as cut-and-count or matrix-based unfolding) increasingly being superseded by more flexible and powerful machine-learning algorithms. We present initial results from applying an unbinned unfolding technique to the WWbb analysis in the lepton+jets channel, highlight its potential advantages over standard unfolding methods, and discuss its implications for Run-III data analyses.

Keywords: Unfolding; Machine Learning; Unbinned