Learning Interpretable Features in Audio Latent Spaces via Sparse Autoencoders

Published in Mechanistic Interpretability Workshop at NeurIPS 2025, 2025