{
  "abstract": "Evaluating a generative model on all queries from all available benchmark tasks is prohibitively expensive. As such, models are often evaluated on different subsets of tasks and/or different subsets of queries from a given task. The resulting collection of responses and the corresponding (model, task) score matrix are sparse and unpaired. In this paper, we introduce the Product Kernel Perspective Space (PKPS), a collection of low-dimensional representations of models based on their embedded responses. PKPS generalizes existing methods for representing black-box generative models by comparing responses to similar -- rather than just identical -- queries and is therefore able to utilize sparse and unpaired information. Theoretically, we show that benchmark prediction using PKPS is query-efficient relative to methods that only use information from dense and paired portions of the evaluation data. Empirically, we show that PKPS-based methods outperform baselines on benchmark prediction across two large benchmark suites.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2611.03614",
  "author_total": 4,
  "authors": [
    {
      "name": "Hayden Helm"
    },
    {
      "name": "Avijit Ghosh"
    },
    {
      "name": "Ben Johnson"
    },
    {
      "name": "Carey Priebe"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/d20sr4f8012jd.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2611.03614",
  "fulltext_url": null,
  "id": "d20sr4f8012jd",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2611.03614",
  "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.LG"
    ],
    [
      "arxiv",
      "stat.ME"
    ],
    [
      "arxiv",
      "stat.ML"
    ]
  ],
  "openalex": null,
  "pdf_url": "https://arxiv.org/pdf/2611.03614",
  "pmcid": null,
  "pmid": null,
  "published": "2026-11-02",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/d20sr4f8012jd.ris",
  "sources": [
    {
      "first_seen": "2026-11-03T12:21:21Z",
      "native_id": "2611.03614",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2611.03614"
    }
  ],
  "subjects": [
    {
      "field": "mathematics",
      "origin": "arxiv:stat.ME",
      "subfield": "statistics"
    }
  ],
  "title": "Benchmark Prediction with Sparse and Unpaired Cached Responses",
  "title_text": "Benchmark Prediction with Sparse and Unpaired Cached Responses",
  "type": "preprint",
  "updated_at": "2026-11-03T12:21:21Z",
  "url": "https://research.draconitic.com/paper/d20sr4f8012jd",
  "venue": "arXiv",
  "versions": []
}