arXiv cs.AI· Ruoran Xu, Wending Gao, Liyunfeng Chen, Aixin Shi, Haoyu Cheng, Zixiang Fang, Yiqiang Zou, Qiufeng Wang·· 3 天前AI 评分35
PhysElite：LLM 距离解决奥赛级物理题还有多远？
PhysElite: How Far Are LLMs from Solving Olympiad-Level Physics Problems?
AI 导读
研究团队发布 PhysElite，一个面向奥赛级物理推理的大规模双语多模态基准，包含 11,586 道奥赛级题目，每题配有图示、中英双语分步解答推导和最终答案。对 18 个开源与闭源 MLLM 的评测显示，最强模型答案准确率仅 33.7%，研究还通过步骤级过程评估诊断模型推理链中的失败环节。数据集已公开。
正文 · 原文
Abstract:Understanding how (multimodal) large language models perform on physics problems requires benchmarks that reflect the difficulty and breadth of expert-level physical reasoning. Existing physics benchmarks remain limited in the following two important ways: (1) short of high-difficulty datasets, and (2) lack of comprehensive coverage of visual forms, knowledge points, and step-by-step solution processes. As a result, model performance on current datasets may not be fully representative of their ability to solve complex physics problems. To address these issues, we present PhysElite, a large-scale bilingual multimodal benchmark for Olympiad-level physics reasoning. PhysElite contains 11,586 Olympiad-tier problems. For each problem, we provide corresponding visual diagrams, step-by-step bilingual Chinese-English solution derivations, and the final answer. We benchmark 18 open-source and closed-source MLLMs, and find that even the strongest model reaches only 33.7% answer accuracy. We additionally conduct step-level process evaluation to diagnose where models fail in the reasoning chain. Our datasets are released at this https URL.
| Comments: | Annual Conference on Neural Information Processing Systems (NeurIPS) 2026 |
| Subjects: | Artificial Intelligence (cs.AI); Multimedia (cs.MM); Mathematical Physics (math-ph) |
| Cite as: | arXiv:2608.25097 [cs.AI] |
| (or arXiv:2608.25097v3 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2608.25097 arXiv-issued DOI via DataCite |
Submission history
From: Ruoran Xu [view email]
[v1]
Tue, 25 Aug 2026 19:48:42 UTC (12,681 KB)
[v2]
Fri, 25 Sep 2026 05:03:30 UTC (12,675 KB)
[v3]
Mon, 12 Oct 2026 23:12:54 UTC (12,676 KB)
来源：arXiv cs.AI · arxiv.org