Startup's new mechanistic interpretability tool lets you debug LLMs

(technologyreview.com)

3 points | by joozio 11 hours ago ago

2 comments