▲ 3 pointsShow HN: HyperSAE – Hyperbolic Sparse Autoencoders for LLM Interpretabilitygithub.comby visha1v·5d ago·0 comments·view on hn ↗