smaddox

1,728 karmajoined January 15, 2015On HN

Recent submissions

Scaling Monosemanticity: Extracting Interpretable Features from Claude 3 Sonnet transformer-circuits.pub
1smaddox2y1 comment
The Matrix: A Bayesian learning model for LLMs arxiv.org
3smaddox2ydiscuss

All submissions and comments on Hacker News.