GPT-6-Astra在理论物理基准TheoPhysEval上解出49/50题
原标题：GPT-6-Astra solves 49 of 50 physics research problems at $2.40 apiece
AI 摘要
来自ICTS、哈佛、KU Leuven和Google DeepMind等机构的35位研究者发布TheoPhysEval基准，包含50道未发表的理论物理研究题。GPT-6-Astra在OpenAI Codex环境中以98%正确率（49/50）居首，平均每题成本不到3美元；Anthropic Opus 5.5得92%，Gemini 3.8 Flash得68%，开源模型DeepSeek-V4.1-Flash和GLM 5.3分别为54%和42%。研究者指出，当前最强模型在具体计算上或已超越顶尖人类研究者，但缺乏提出正确问题和判断问题价值的“品味”，因此尚未带来理论物理突破。
正文节选
GPT-6-Astra solves 49 of 50 physics research problems at $2.40 apiece A new benchmark shows the best AI agents can solve theoretical physics calculations as accurately as leading specialists in each subfield. The researchers behind it say a missing sense of "taste" is one reason breakthroughs haven't followed yet. Thirty-five researchers from institutions including the International Centre for Theoretical Sciences (ICTS), Harvard University, KU Leuven, and Google DeepMind have released TheoPhysE