[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"news-e2eb81dc-3112-411a-94b0-b5061a12be78":3},{"id":4,"title":5,"summary":6,"original_url":7,"source_id":8,"tags":9,"published_at":23,"created_at":24,"modified_at":25,"is_published":26,"publish_type":27,"image_url":13,"view_count":28},"e2eb81dc-3112-411a-94b0-b5061a12be78","AdvancedMathBench 把数学证明拉进博士级:GPT-5.5-xhigh 仍只 75.8","当大多数 benchmark 还在用高中\u002F奥数级数学题考 LLM 时,Intern Large Models(上海 AI Lab)推出的 AdvancedMathBench 把题目难度直接拉到「本科生高年级(UGD)+ 博士资格考(QE)」级别——核心 ProverBench 收录 296 道这样的难题,配套 VerifierBench 又用 888 条模型生成的证明轨迹去测「验证能力」。\n\n实验结果相当难看。在证明生成上,目前最强的 GPT-5.5-xhigh 也只拿下 75.8(UGD)和 66.1(QE),意味着即使是最顶级的模型,在博士级数学证明上也有 1\u002F4 到 1\u002F3 的题目彻底做不出来。在证明验证上,最强模型 Balanced F1 只有 65.1,而且所有模型的 True Negative 普遍偏低——模型抓「错误证明」的本事远远不够,容易把错的当成对的过。\n\n这套 benchmark 的最大贡献,不在于「又一次证明 LLM 不会做难题」,而在于把评估颗粒度从「最终答案对错」细化到「证明过程是否有效」,用 verifier pipeline + 专家标注做 fine-grained 错误定位。对 agentic 工作流(让 LLM 互相审稿、互相改稿)非常关键——如果 verifier 抓不到漏洞,整个 agent 链条的可信度就是空话。\n\n对比 OpenAI 复审 SWE-Bench Pro 时承认有近三成「坏题」,AdvancedMathBench 选择了更难的方向:题目可能更干净,但评判标准更严苛。这或许暗示 LLM 评估正在从「刷榜」走向「过程审计」的下一阶段。","https:\u002F\u002Farxiv.org\u002Fabs\u002F2607.11849","7437aeb9-930c-4866-a2e9-48003c1a792b",[10,14,17,20],{"id":11,"name":12,"slug":12,"description":13,"color":13},"5e628969-6d2a-437f-998a-104e4b16cfb1","ai-progress",null,{"id":15,"name":16,"slug":16,"description":13,"color":13},"120fa59a-ff6f-4537-9bf5-f818df636a0e","benchmark",{"id":18,"name":19,"slug":19,"description":13,"color":13},"0a93ec8e-ea39-4693-81de-563ca8c173f7","inference",{"id":21,"name":22,"slug":22,"description":13,"color":13},"01598627-1ea6-4b27-a5d8-874971571a71","llm","2026-07-14T16:15:00Z","2026-07-14T16:12:26.556700Z","2026-07-14T16:12:26.556711Z",true,"agent",2]