Emmanuel Ameisen
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
Exploring the Biology of LLMs with Circuit Tracing with Emmanuel Ameisen
- Published
- April 14, 2025
- Duration
- 1h 34m
- Summary source
- description
- Last updated
- Jun 7, 2026
Discusses llm, anthropic, neural-nets.
Summary
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 …
Intelligent report
Sign in to read teasers, or upgrade to Research Pro to commission a new dossier for this episode. Learn more →
Show notes
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 sparse, interpretable alternatives. The conversation explores several fascinating discoveries about large
Themes
- llm
- anthropic
- neural-nets