Goodfire's interpretability breakthrough splits OpenAI researchers' takes
burny_tech · x · 2026-10-17
Dean Ball congratulates Nick and Goodfire, calling it possibly one of the most important single days in deep learning history. But observer evzenwy notes a split timeline: half of OpenAI non-interpretability staff hail it as the second coming of Christ, while interpretability researchers (including Leo Gao) say it's... fine.
More from Companies & People
- Open textbook "Elements of Robotics" by Ben-Ari & Mondada now freely available — burkov · 2026-10-17
- VCs underwrite other investors' interest, not companies, as capital glut distorts exits — arian_ghashghai · 2026-10-17
- Fresh Instagram account shows sexualized search results with zero interaction history — NathanpmYoung · 2026-10-17
- Meta drops 816B open-weight MoE, Anthropic hits $16.8B Q3, 8 states sue OpenAI — rohanpaul_ai · 2026-10-17
- Beff Jezos says his brain hit token limits: 16-hour days, 80-hour weeks — beffjezos · 2026-10-17
- Replit CEO says it stopped working with Anthropic, citing cost and model tests — ryanmerket · 2026-10-17