Speaker
Description
This poster presents a study on the tau ($\tau$) identification efficiency in $t\bar{t} \rightarrow \ell\tau_{\text{had}} \nu\bar{\nu} b\bar{b}$ events, using 29 fb$^{-1}$ of proton-proton collision data at $\sqrt{s} = 13.6$ TeV, recorded by the ATLAS detector during LHC Run-3. A tag-and-probe method is applied to measure the efficiency of the RNN-based tau-ID algorithm, focusing on hadronically decaying tau leptons ($\tau_{\text{had}}$) in events enriched with top quark-antiquark pairs.
Key distributions, such as transverse momentum ($p_T$) and pseudorapidity ($\eta$) of $\tau_{\text{had}}$, along with RNN scores, are analyzed to evaluate the algorithm’s performance. Scale factors for $\tau$-ID are extracted through simultaneous fitting, incorporating systematic uncertainties and nuisance parameters. The results provide insight into tau-ID performance in a challenging $t\bar{t}$ environment, offering valuable contributions to precision modeling of tau decay processes in the ATLAS experiment.
