{
  "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 https://huggingface.co/datasets/physelite/PhysElite.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2608.25097",
  "author_total": 8,
  "authors": [
    {
      "name": "Ruoran Xu"
    },
    {
      "name": "Wending Gao"
    },
    {
      "name": "Liyunfeng Chen"
    },
    {
      "name": "Aixin Shi"
    },
    {
      "name": "Haoyu Cheng"
    },
    {
      "name": "Zixiang Fang"
    },
    {
      "name": "Yiqiang Zou"
    },
    {
      "name": "Qiufeng Wang"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/e02xcr0be93dq.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2608.25097",
  "fulltext_url": null,
  "id": "e02xcr0be93dq",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2608.25097",
  "language": null,
  "licence": "CC-BY-4.0",
  "licence_label": "CC BY 4.0",
  "licence_source": "article",
  "licence_url": "http://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": "CC0-1.0",
  "native_subjects": [
    [
      "arxiv",
      "cs.AI"
    ],
    [
      "arxiv",
      "cs.MM"
    ],
    [
      "arxiv",
      "math-ph"
    ],
    [
      "arxiv",
      "math.MP"
    ],
    [
      "openalex",
      "1707|Multimodal Machine Learning Applications"
    ]
  ],
  "openalex": "W7204442925",
  "pdf_url": "https://arxiv.org/pdf/2608.25097",
  "pmcid": null,
  "pmid": null,
  "published": "2026-10-12",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/e02xcr0be93dq.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:29:11Z",
      "native_id": "2608.25097",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2608.25097"
    }
  ],
  "subjects": [
    {
      "field": "mathematics",
      "origin": "arxiv:math-ph",
      "subfield": "mathematical-physics"
    }
  ],
  "title": "PhysElite: How Far Are LLMs from Solving Olympiad-Level Physics Problems?",
  "title_text": "PhysElite: How Far Are LLMs from Solving Olympiad-Level Physics Problems?",
  "type": "preprint",
  "updated_at": "2026-10-30T12:29:19Z",
  "url": "https://research.draconitic.com/paper/e02xcr0be93dq",
  "venue": "arXiv",
  "versions": []
}