We used sparse autoencoders to explain LLM moderation flags of violent threats (variance.co)

6 points by karinemellata 1y ago ↗ HN

0 comments

[ 3.0 ms ] story [ 7.6 ms ] thread

No comments yet.