ICLR hits 60,000 submissions as MIT professor warns peer review is broken
CSProfKGD · x · 2026-09-22
MIT professor Phillip Torr flags ICLR's 60,000 submissions and questions what it means for AI and science: ML reviews were always noisy, good reviewers are now impossible to find, and the reciprocal review rule is "badly broken." Oral exams won't scale, and AI reviewing would just mean AI-generated papers reviewed by AIs. As ICCV'29 General Chair, he's asking for solutions.
Related event: ICLR 2027 tops 60k submissions, sparking calls for peer-review reform(8 posts)→
More from Research
- 200-seat UC Berkeley course on full-stack post-training, RL and applied RSI announced — karinanguyen · 2026-11-12
- NVIDIA's Déjà View: 117M looped transformer matches models 10x larger at 3D reconstruction — NVIDIAAI · 2026-11-12
- OpenAI finds models are getting harder to monitor through their chain-of-thought — xiamengzhou · 2026-11-12
- Show Reddit: ge, a Rust GGUF engine, benchmarks against llama.cpp on a 4GB RTX 3050 — Raptor2496 · 2026-11-12
- 1,500 AI-for-Science Papers in a Week Share One Question: Compared With What? — bravo_abad · 2026-11-12
- AI system proves longstanding quantum computing conjecture in three days, Lean-verified — _arohan_ · 2026-11-12