Philippe Chlenski

@philippe-chlenski · 1 works

Researches mechanistic interpretability methods for language models without full internal access.

Recent posts by Philippe Chlenski

• Surrogate Fidelity: When Can Open LLMs Explain Closed Ones? - Mechanistic interpretability (MI) requires full access to model internals, yet the APIs for most widely deployed language models at best expose log-probabilities over output tokens. This creates a… · Paper · arxiv.org