Preliminary Studies of ttH(bb) Dileptonic Events Using CMS Run3 Data and Initial Tests of SPANet in This Channel

The Standard Model (SM) of particle physics predicts that the Higgs boson couples to fermions via a Yukawa-type interaction, with a strength proportional to the fermion mass. This makes the associated production of a Higgs boson with a top-quark pair ( ttH ) a crucial process to directly probe the top-Higgs Yukawa coupling, an essential parameter for confirming the SM nature of the Higgs boson. Among Higgs boson decays, the channel into a bb quark pair has the largest branching fraction, offering an experimentally promising final state. However, ttH(bb) process faces significant challenges regarding backgrounds, especially tt +jets production, with the ttbb background being irreducible with respect to the ttH, H → bb signal. Advance Machine Learning techniques are essential to improve the sensitivity to the signal process.This work focuses on the analysis of the ttH, H → bb process in events with two leptons, using proton-proton collision data collected by the CMS experiment at the LHC during Run3 at √s = 13.6 TeV . ML methods are explored to significantly enhance the sensitivity to the ttH signal. For the first time in this final state we are exploring the potential of SPANet for jet-parton assignment and neutrino kinematic regressions and finally signal and background classification.