Sparse Auto-Encoders (SAEs) for Mechanistic Interpretability Collection A compilation of sparse auto-encoders trained on large language models. • 37 items • Updated 13 days ago • 16