{
  "abstract": "Benchmark contamination, the leakage of test items into training data, is widely described as a threat to the reliability of large language model (LLM) leaderboards. We argue that this concern conflates two distinct questions: whether contamination inflates absolute scores, and whether it reorders the ranking of models. We recast contamination as a violation of anchor-item invariance and measure it through the differential functioning of original versus semantically equivalent paraphrased items, a within-item contrast that holds the measured skill fixed and isolates memorization from capability. Using per-instance responses from 47 publicly released models and 74 models finetuned with a known dose of contamination, across four benchmarks (ARC, GSM8K, HellaSwag, MMLU), we first calibrate the measure against ground truth: it recovers injected contamination dose-responsively (a corrected effect of +0.187 accuracy points for test-set leakage) and never flags a negative-control model trained only on the legitimate training split (-0.012). We then quantify leaderboard impact: the rank correlation between a standard leaderboard and a paraphrase-controlled leaderboard is 0.997, and a sensitivity analysis shows that the observed differential contamination is far below the level needed to move rankings, with only 3 of 188 model-by-benchmark cases showing differential contamination corroborated across two references. Contamination among these public models is therefore largely uniform: it inflates absolute scores without reordering the leaderboard, and ranking distortion requires the rare case of differential contamination. We provide a calibrated invariance audit, released as a reference implementation, and recommend that leaderboards report paraphrase-controlled rankings alongside confidence intervals.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2609.02899",
  "author_total": 2,
  "authors": [
    {
      "name": "Xingyao Xiao"
    },
    {
      "name": "Yihong Cheng"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/0qbssca7e9tg2.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2609.02899",
  "fulltext_url": null,
  "id": "0qbssca7e9tg2",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2609.02899",
  "language": null,
  "licence": "LicenseRef-arXiv-nonexclusive-1.0",
  "licence_label": "arXiv licence",
  "licence_source": "article",
  "licence_url": "http://arxiv.org/licenses/nonexclusive-distrib/1.0/",
  "metadata_licence": "CC0-1.0",
  "native_subjects": [
    [
      "arxiv",
      "cs.CL"
    ],
    [
      "arxiv",
      "stat.AP"
    ],
    [
      "arxiv",
      "stat.ME"
    ],
    [
      "openalex",
      "1702|Topic Modeling"
    ]
  ],
  "openalex": "W7208759730",
  "pdf_url": "https://arxiv.org/pdf/2609.02899",
  "pmcid": null,
  "pmid": null,
  "published": "2026-09-01",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/0qbssca7e9tg2.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:18:42Z",
      "native_id": "2609.02899",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2609.02899"
    }
  ],
  "subjects": [
    {
      "field": "mathematics",
      "origin": "arxiv:stat.AP",
      "subfield": "statistics"
    }
  ],
  "title": "Contamination Inflates Scores but Rarely Reorders Large Language Model Leaderboards",
  "title_text": "Contamination Inflates Scores but Rarely Reorders Large Language Model Leaderboards",
  "type": "preprint",
  "updated_at": "2026-10-30T12:18:42Z",
  "url": "https://research.draconitic.com/paper/0qbssca7e9tg2",
  "venue": "arXiv",
  "versions": []
}