Mechanistic Interpretability Benchmark

AI & ML interests

Principled evaluation of mechanistic interpretability methods.

Recent Activity

atticusg  authored a paper 8 days ago

Open Problems in Mechanistic Interpretability

hadasor  authored a paper 22 days ago

Padding Tone: A Mechanistic Analysis of Padding Tokens in T2I Models

hadasor  authored a paper 4 months ago

LLMs Know More Than They Show: On the Intrinsic Representation of LLM Hallucinations

View all activity

Team members 17