GPT-6 Astra Only Non-Zero Scorer on 68 Unsolved Erdős Problems, Solves 5
量子位 · wechat · 2026-09-04
Epoch AI and the University of Manchester launched FrontierMath Erdős (FME), a benchmark of 68 unsolved Erdős problems requiring Lean-formalized proofs — a direct response to Terence Tao's critique of uncontrolled AI math claims. GPT-6 Astra was the only model to score non-zero (3%); with relaxed budgets it solved 5 conjectures total, including Erdős's 1931 problem #1 and the Erdős-Sós conjecture. All attempts cost over $220K, implying $10K expected cost per solved open problem; 63 problems remain unsolved.
Related event: Epoch AI Launches FrontierMath Erdős; GPT-6 Astra First to Break Zero(10 posts)→
More from Models
- Opus 5.5 tops OpenRouter spending in 20 of 29 task categories — mmurph · 2026-10-21
- Meta team builds open-source 30B LLM from scratch in months, hails 'second era' — TimDarcet · 2026-10-21
- DFlash drafter for Muse Glimmer-30B: 5 layers, blockwise 16-token spec decoding — TimDarcet · 2026-10-21
- EvoPress quantization shrinks 30B model from 56GB to 19GB at near-bf16 accuracy — TimDarcet · 2026-10-21
- Radar chart: Muse Glimmer-30B benchmarked against Gemma4-31B and Qwen3.6-27B — TimDarcet · 2026-10-21
- Privacy evals: Muse Glimmer-30B leaks less than Qwen3.6-27B on CIMemories — TimDarcet · 2026-10-21