We're sunsetting PodQuest on 2025-07-28. Thank you for your support!
Export Podcast Subscriptions
back
“Scaling Sparse Feature Circuit Finding to Gemma 9B” by Diego Caples, Jatin Nainani, CallumMcDougall, rrenaud
33:26
Share
2025/1/10
LessWrong (30+ Karma)
AI Chapters
Transcribe
Chapters
What's the TL;DR of the Research?
Introduction to the Research
Background on SAEs and Circuits
What Are the Problems with Current Sparse Feature Interpretability Approaches?
How Does Our Approach Solve Scalability and Independent Scoring Issues?
What Are the Key Results of the Research?
How Stable Are the Masks in the Circuit Finding Algorithm?
Case Study: Code Output Prediction
What Are the Conclusions of the Research?
What Future Research and Ideas Are on the Horizon?
Shownotes
Transcript
No transcript made for this episode yet, you may request it for free.