{
  "abstract": "Evo 2 is the largest openly available genomic foundation model, but its forty billion parameter configuration cannot be loaded onto a single 80 GB accelerator, placing genome-scale analysis beyond most laboratories. We present TurboQuant-Bio, an open toolkit that compresses Evo 2s weights and attention cache to four bits without calibration data, and serves both through fused kernels. Compression is near-lossless across perplexity spanning the tree of life, genomic classification, splice-site prediction, gene completion and clinically relevant variant-effect prediction. It brings Evo 2 40B onto one 80 GB GPU and Evo 2 7B to its full million-token context within a 40 GB memory budget, an eightfold gain in reachable context. We further show that the released chunked-prefill path is silently incorrect, returning plausible but uncorrelated likelihoods, and derive the block-wise continuation that repairs it: a complete 580-kilobase bacterial genome is now scored in one context in 22 minutes rather than 13.7 hours.",
  "abstract_shown": true,
  "affiliations": [
    "The University of Texas at Austin"
  ],
  "article_type": "New results",
  "arxiv": null,
  "author_total": 3,
  "authors": [
    {
      "name": "M. Patsakis"
    },
    {
      "name": "A. Tzanakakis"
    },
    {
      "affs": [
        0
      ],
      "name": "I. Georgakopoulos-Soares"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/7ap9d96f83b7c.bib",
  "cited_by": 0,
  "doi": "10.64898/2026.08.28.747902",
  "fulltext_url": "https://www.biorxiv.org/content/10.64898/2026.08.28.747902v1.full",
  "id": "7ap9d96f83b7c",
  "issns": [],
  "landing_url": "https://www.biorxiv.org/content/10.64898/2026.08.28.747902v1",
  "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",
      "bioinformatics"
    ],
    [
      "openalex",
      "1707|Advanced Neural Network Applications"
    ]
  ],
  "openalex": "W7204638432",
  "pdf_url": "https://www.biorxiv.org/content/10.64898/2026.08.28.747902v1.full.pdf",
  "pmcid": null,
  "pmid": null,
  "published": "2026-08-28",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/7ap9d96f83b7c.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:18:13Z",
      "native_id": "10.64898/2026.08.28.747902",
      "source": "biorxiv",
      "url": "https://www.biorxiv.org/content/10.64898/2026.08.28.747902v1"
    }
  ],
  "subjects": [
    {
      "field": "biology",
      "origin": "biorxiv:bioinformatics",
      "subfield": "bioinformatics"
    }
  ],
  "title": "Calibration-free compression brings Evo 2 to its full million-token context on a single GPU",
  "title_text": "Calibration-free compression brings Evo 2 to its full million-token context on a single GPU",
  "type": "preprint",
  "updated_at": "2026-10-30T12:18:13Z",
  "url": "https://research.draconitic.com/paper/7ap9d96f83b7c",
  "venue": "bioRxiv",
  "versions": []
}