Episode
Exploring the Biology of LLMs with Circuit Tracing with Emmanuel Ameisen - #727
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence) · Apr 14, 2025 · 1 h 34 min
In this episode, Emmanuel Ameisen, a research engineer at Anthropic, returns to discuss two recent papers: "Circuit Tracing: Revealing Language Model Computational Graphs" and "On the Biology of a Large Language Model." Emmanuel explains how his team developed mechanistic interpretability methods to understand the internal workings of Claude by replacing dense neural network components with…