Separating instrumental glitches from generic signals in gravitational-wave data

Transient noise artifacts (called ``glitches'') frequently occur in ground-based gravitational-wave detectors, sometimes in close temporal coincidence to real GW events. As more events are detected, mitigation of these glitches becomes increasingly important. Previously, the BayesWave algorithm has shown success in removing glitches from data used for binary coalescence parameter estimation. As we observe more of the gravitational-wave universe, we may discover signals from novel sources lacking robust models. In such cases, confidently distinguishing generic transients from glitches will be crucial. This work investigates BayesWave's ability to recover simulated ad-hoc gravitational-wave signals overlapping instrumental glitches. We will present results on the match between recovered and simulated waveforms, and how well the algorithm correctly identifies data containing both signals and glitches.