OpenAI's 11-day-trained model and ~10,000 coordinating agents cracked Navier–Stokes
peterwildeford · x · 2026-09-09
Reacting to OpenAI's technical disclosure of the Navier–Stokes proof, peterwildeford marvels that a model trained for only 11 days was useful on a Millennium Problem.
Details from OpenAI: a system of coordinating agents powered by their internal model, with access to a cached internet snapshot and code execution; agents were subdivided into communicating groups of varying sizes, and the group producing the Navier–Stokes resolution involved on the order of 10,000 agents.
More from Models
- DeepSeek v4 and GLM Now Run Faster Than vLLM and SGLang — jedisct1 · 2026-09-09
- OpenAI Claims Agent-Generated Proof of Navier-Stokes Millennium Problem — soumitrashukla9 · 2026-09-09
- Magic details >10x compute-efficient pretraining, eyes trillion-parameter models — Dr_Singularity · 2026-09-09
- Magic claims >10x more efficient pretraining, matches DeepSeek V4 Pro with 50x fewer FLOPs — Dr_Singularity · 2026-09-09
- User claims running 'gpt-6-astra' for over an hour at just $7.06 — arthurcolle · 2026-09-09
- Qwen3.8 27B quantization benchmark: 4-bit holds up, 1-bit collapses — victormustar · 2026-09-09