Interpretability-guided weight edits for Gemma 270M, no training
Comments
Ranked by a public rubric, not by votes: what a comment adds counts, not who wrote it. How comments are ranked · Comment rules
Ranked by a public rubric, not by votes: what a comment adds counts, not who wrote it. How comments are ranked · Comment rules
Summary
A Hacker News post by an individual developer describes making Google's 270-million-parameter Gemma 3 model better at following instructions without any gradient-based training, instead editing a few weights directly.