{
  "oags": "0.2",
  "status": "alpha",
  "entry": {
    "id": "IB-arxiv-2606.06117-pvr-vs-nucleotide-transformer",
    "type": "evidence",
    "claim": "On three low-sample genomic benchmarks, pVR outperformed zero-shot frozen embeddings from the 500M-parameter Nucleotide Transformer v2 by 6.7 to 11.4 percentage points.",
    "evidence_grade": "E3",
    "state": "active",
    "primary_source_ref": [
      {
        "kind": "arxiv",
        "id": "2606.06117",
        "version": "v1",
        "url": "https://arxiv.org/abs/2606.06117",
        "locator": "abstract"
      }
    ],
    "quantities": {
      "gain_pp_low": 6.7,
      "gain_pp_high": 11.4,
      "nt_params_m": 500
    },
    "declared_loss": null,
    "source_title": "p-adic Bi-Filtrations for Topological Machine Learning on Genomic Sequences",
    "authors": [
      "Tirtharaj Dash",
      "Gunja Sachdeva"
    ],
    "result_origin": "measured",
    "support_quote": "pVR also outperforms zero-shot frozen embeddings from the 500M-parameter Nucleotide Transformer v2 by $6.7$ to $11.4$ percentage points on three low-sample benchmarks",
    "source_version": "v1",
    "abstract_check": {
      "version": "v1",
      "date": "2026-10-08",
      "quote_verbatim": true
    },
    "domain_v2": "q-bio/genomics"
  },
  "expand": [],
  "incoming": [],
  "links": [],
  "declared_losses": [
    "Public evidence-graph projection: internal analytic/theory edges and private provenance are withheld by design.",
    "Alpha: some primary-source refs await precise citation verification."
  ],
  "generated": "2026-10-08"
}