13 LLMs with 320 past experiments in context all beat baseline at predicting research payoff
cihangxie · x · 2026-11-11
Can LLMs build a 'Research World Model'? Researchers asked 13 LLMs to predict each research idea's gain before running it, with 320 past experiments in context. All 13 beat the best model given none, suggesting LLMs can acquire a form of research taste and be used to triage experiment proposals.
Related event: Research World Models: LLMs Predict Which Experiments Are Worth Running(2 posts)→
More from AGI Musings
- METR report: AI-driven R&D acceleration puts recursive self-improvement within reach by 2027 — birchlse · 2026-11-11
- Researchers launch non-profit STABL to study AI's societal bottlenecks — sayashk · 2026-11-11
- "Retrofitting electricity to coal-powered factories": one line on AI's institutional mismatch — curious_vii · 2026-11-11
- 'AI can't be close to AGI, cancer isn't cured' — the dung beetle fallacy — danfaggella · 2026-11-11
- Yann LeCun: 'If you are interested in human-level AI, don't work on LLMs' — StasBekman · 2026-11-11
- US robotaxi deployments are closing the gap with China, new data shows — skorusARK · 2026-11-11