OneBench
On the Interpretability of Whisper Encodings Using Sparse Autoencoders | OneBench: AI Insights