Show HN: HyperSAE – Hyperbolic Sparse Autoencoders for LLM Interpretability (github.com)

3 points by visha1v 2mo ago ↗ HN

0 comments

[ 3.3 ms ] story [ 2.6 ms ] thread

No comments yet.

A showcase of shard-db. Data from the Hacker News API. Refreshed every 15 minutes.

Source · Live DB stats