Interpretability researcher maxsloef leaves Goodfire to study layered persona formation in LLMs
Interpretability researcher maxsloef has left Goodfire to study how model personas form across layers. He supports a layered view of LLMs—early layers translate tokens into concepts, middle layers compute over them, and later layers translate back—though which layers determine persona remains an open question.
2026-11-01 ~ 2026-11-01 · 2 related posts
- Researchers debate LLM layerwise picture: where does a model's persona live? — maxsloef · 2026-11-01
- Interpretability researcher maxsloef leaves Goodfire to study how LLM personas form — maxsloef · 2026-11-01