Show HN: HyperSAE – Hyperbolic Sparse Autoencoders for LLM Interpretability

 (github.com)

3 points | by visha1v 6 hours ago ago

No comments yet.