{
  "abstract": "A common assumption in language model development is that cognitive abilities are organized around a general, domain-free intelligence factor, like fluid intelligence in humans. This assumption is rarely tested directly, and prior attempts have done so only at a much smaller scale. We take a latent variable approach to intelligence in language models, similar to how psychometricians study psychological constructs. Performance in every specific problem set is influenced by a domain-specific and a domain-agnostic latent factor. Using factor analysis as a dimension-reduction technique, we analyzed 13,251 published evaluation scores covering 1,618 language models across 456 different text-only benchmarks. Due to the super-sparse nature of the dataset, we triangulate our analysis across different data densifiers and imputation methods. A robust pattern across different modes of bias is that 1. A general intelligence factor accounts for 70.8% of variance in model performance at our most generous estimate, and far less than that in most of our solutions, 2. Content-similar benchmarks do not necessarily cluster together, and 3. The $g$ factor is not dominated by any common theme, and there is a lack of evidence that it is well-proxied by standard \"intelligence\" benchmarks. Our findings go against current endeavors of defining, identifying, and targeting general intelligence as a tangible construct in language model development. This leaves the strategy of targeting a single conceptual ability without support, since the first-order abilities it would have to reach are often partially idiosyncratic and not identifiable in practice.",
  "abstract_shown": true,
  "affiliations": [],
  "article_type": null,
  "arxiv": "2609.36515",
  "author_total": 3,
  "authors": [
    {
      "name": "Faiz Ghifari Haznitrama"
    },
    {
      "name": "Afrizal Hasbi Azizy"
    },
    {
      "name": "Faeyza Rishad Ardi"
    }
  ],
  "bibtex": "https://research.draconitic.com/paper/bxaadbr1x8jb6.bib",
  "cited_by": 0,
  "doi": "10.48550/arxiv.2609.36515",
  "fulltext_url": null,
  "id": "bxaadbr1x8jb6",
  "issns": [],
  "landing_url": "https://arxiv.org/abs/2609.36515",
  "language": null,
  "licence": "CC-BY-NC-SA-4.0",
  "licence_label": "CC BY-NC-SA 4.0",
  "licence_source": "article",
  "licence_url": "http://creativecommons.org/licenses/by-nc-sa/4.0/",
  "metadata_licence": "CC0-1.0",
  "native_subjects": [
    [
      "arxiv",
      "cs.AI"
    ],
    [
      "arxiv",
      "cs.CL"
    ],
    [
      "arxiv",
      "q-bio.NC"
    ],
    [
      "openalex",
      "1702|Topic Modeling"
    ]
  ],
  "openalex": "W7214937092",
  "pdf_url": "https://arxiv.org/pdf/2609.36515",
  "pmcid": null,
  "pmid": null,
  "published": "2026-09-29",
  "publisher": null,
  "ris": "https://research.draconitic.com/paper/bxaadbr1x8jb6.ris",
  "sources": [
    {
      "first_seen": "2026-10-30T12:25:36Z",
      "native_id": "2609.36515",
      "source": "arxiv",
      "url": "https://arxiv.org/abs/2609.36515"
    }
  ],
  "subjects": [
    {
      "field": "biology",
      "origin": "arxiv:q-bio.NC",
      "subfield": "neuroscience"
    }
  ],
  "title": "Large-scale factor analysis shows machine intelligence is only partially interpretable",
  "title_text": "Large-scale factor analysis shows machine intelligence is only partially interpretable",
  "type": "preprint",
  "updated_at": "2026-10-30T12:25:36Z",
  "url": "https://research.draconitic.com/paper/bxaadbr1x8jb6",
  "venue": "arXiv",
  "versions": []
}