gpt-6POST
Will Depue 推测 OpenAI 新模型是全新预训练而非 RL 微调
他根据 OpenAI 员工反复暗示判断这是新预训练模型，称其预训练不到两周就超过 Astra 并解出 Navier-Stokes，直呼疯狂。
will depue
@willdepue2026.09.16收录原推发布于 2026.09.15
will depue @willdepue原推发布于 2026.09.15
openai people keep gesticulating towards the ‘new model‘ so much so that i’m absolutely sure it’s a new pretrain and not some RL run, which implies (i think) OpenAI has a model that not more than 2 weeks into pretraining was better than Astra and solved Navier-Stokes. insane
这条是引用着另一条写的：x.com/polynoamial/status/2099726370356314563
x.com/willdepue/status/2099746385486635469采集于 2026.09.16 10:17