{
  "abstract": "Inferring biological function from experimental data is central to understanding emerging pathogens and developing effective countermeasures, yet interpreting these data remains slow and expert-intensive. AI agents could help accelerate this process by reasoning across sequence, structural, and biophysical evidence. We present BioSecBench-Function, a verifiable benchmark for recovering biosecurity-relevant function from real biological data. The benchmark comprises 111 evaluations built from published datasets and graded deterministically against ground truth. We organize evaluations along two dimensions: threat axis (spanning seven biosecurity-relevant question types) and biological question (indicating whether the solution depends primarily on sequence, structure, or biophysical assay data). Across 7,326 runs from twenty-two model-harness configurations, Opus 5 under Claude Code led on endpoint pass rate at 50.3%, and Grok 4.6 under Grok Build led on overall pass rate at 44.1% when refusals counted as failures. Performance varied substantially across both model-harness configurations and task categories. Refusal rates differed sharply by provider, and cost was a poor predictor of accuracy: several configurations exceeded 40% pass rate at low cost. BioSecBench-Function provides a standard for measuring whether agents can be trusted to interpret what a new pathogen or variant does when the next outbreak arrives.",
  "abstract_shown": true,
  "affiliations": [
    "LatchBio"
  ],
  "article_type": "New results",
  "arxiv": null,
  "author_total": 11,
  "authors": [
    {
      "affs": [
        0
      ],
      "name": "D. Wang"
    },
    {
      "name": "Q. Xu"
    },
    {
      "name": "A. Banerjee"
    },
    {
      "name": "R. Jain"
    },
    {
      "name": "A. Vermani"
    },
    {
      "name": "J. Felix"
    },
    {
      "name": "G. Moreno"
    },
    {
      "name": "F. AlZaben"
    },
    {
      "name": "N. Keul"
    },
    {
      "name": "E. Seeyave"
    },
    {
      "name": "H. Bhasin"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/7x1j9rdbcgdax.bib",
  "cited_by": 0,
  "doi": "10.64898/2026.09.03.749010",
  "fulltext_url": "https://www.biorxiv.org/content/10.64898/2026.09.03.749010v1.full",
  "id": "7x1j9rdbcgdax",
  "issns": [],
  "landing_url": "https://www.biorxiv.org/content/10.64898/2026.09.03.749010v1",
  "language": null,
  "licence": "CC-BY",
  "licence_label": "CC BY",
  "licence_source": "article",
  "licence_url": "https://creativecommons.org/licenses/by/4.0/",
  "metadata_licence": null,
  "native_subjects": [
    [
      "biorxiv",
      "biophysics"
    ],
    [
      "openalex",
      "1312|Machine Learning in Bioinformatics"
    ]
  ],
  "openalex": "W7211930174",
  "pdf_url": "https://www.biorxiv.org/content/10.64898/2026.09.03.749010v1.full.pdf",
  "pmcid": null,
  "pmid": null,
  "published": "2026-09-03",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/7x1j9rdbcgdax.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:20:18Z",
      "native_id": "10.64898/2026.09.03.749010",
      "source": "biorxiv",
      "url": "https://www.biorxiv.org/content/10.64898/2026.09.03.749010v1"
    }
  ],
  "subjects": [
    {
      "field": "biochemistry",
      "origin": "biorxiv:biophysics",
      "subfield": "biophysics"
    }
  ],
  "title": "BioSecBench-Function: A Verifiable Benchmark for Reasoning about Biological Function from Experimental Data",
  "title_text": "BioSecBench-Function: A Verifiable Benchmark for Reasoning about Biological Function from Experimental Data",
  "type": "preprint",
  "updated_at": "2026-10-30T12:20:18Z",
  "url": "https://research.draconitic.com/paper/7x1j9rdbcgdax",
  "venue": "bioRxiv",
  "versions": []
}