Researchers teach LLMs to find interesting theorems, boosting discovery 4.3x
burny_tech · x · 2026-09-27
A new effort by @niketnpatel and @KempeLab targets the next bottleneck in AI math discovery: while LLMs already prove results that resisted mathematicians for decades, finding interesting theorems without human guidance remains hard. The team introduces a quantitative notion of interestingness, trains an LLM to optimize it (4.3x higher interestingness), and builds a self-expanding discovery loop.
Related event: NYU's Interestingness Metric Boosts AI Math Discovery 4.3x(4 posts)→
More from Research
- 7 Key Studies on LLM-Assisted Peer Review, From Bias to Faulty Reasoning — sethlazar · 2026-09-27
- Tauon optimizer beats Muon on GPT-Mini: lower loss and ~8.5% faster steps — kkkrlklo · 2026-09-27
- Rufus-Air paper: ordering post-training by reward reliability builds competitive open LLM — burny_tech · 2026-09-27
- DeepMind's Economic Policy for AGI framework evaluates 11 interventions — CurieuxExplorer · 2026-09-27
- Training RL Agents to Play a Street Fighter-Like Game Reveals Heavy Reward Hacking — microscope1024 · 2026-09-27
- Contrastive World Models Learns World Models in Latent Space Without Pixel Prediction — burny_tech · 2026-09-27