This study presents the first comprehensive machine-learning follow-up of the gravitational-wave event GW231123, the most massive binary black-hole merger detected to date (190–265 M⊙). Using three complementary AI tools—GW-Whisper (a Whisper-based transformer for signal classification), ArchGEM (a Gaussian-mixture model for glitch characterization), and AWaRe (an attention-based autoencoder for waveform reconstruction)— the work confirms the astrophysical origin of the event, extracts physical parameters of scattered-light glitches, and reconstructs the waveform with high fidelity. The framework demonstrates robust performance across simulated mergers up to 1000 M⊙, providing a scalable, model-agnostic approach for rapid validation of intermediate-mass black-hole mergers in noisy LIGO–Virgo–KAGRA data.
Dr. Chayan Chatterjee, Kaylah McGowan, Suyash Deshmukh, Karan Jani
