{
  "id": "page:field",
  "type": "WebPage",
  "canonical_url": "https://www.honesty.tools/field",
  "source_url": "https://www.honesty.tools/field",
  "title": "Field guide to AI honesty — Honesty.Tools",
  "summary": "Start here. A field guide to AI honesty and safety: the questions researchers ask, the public datasets they test with, the tools they use, and what our audits found when we checked the tests themselves.",
  "status": "Published public page; research and product status remains as stated in the source text.",
  "published_at": null,
  "updated_at": null,
  "last_verified_at": null,
  "date_basis": "No normalized editorial or research-verification date asserted. Dated statements remain in the source text.",
  "verification_basis": "Static authored content from the public HTML. Interactive outputs are not simulated. Source-specific research, product and evidence limits remain in the text; this export is not new verification.",
  "source_revision": "sha256:af52ee3800c5598185e6a9d76909c3b2c14fe421d7826cc90b13bdfd8ff98e66",
  "source_revision_basis": "Public source HTML excluding generated representation links and head metadata; a content revision, not a research verification.",
  "sources": [
    {
      "label": "hallucination",
      "url": "https://www.honesty.tools/field#w-hallucination"
    },
    {
      "label": "TruthfulQA",
      "url": "https://arxiv.org/abs/2109.07958"
    },
    {
      "label": "HaluEval",
      "url": "https://arxiv.org/abs/2305.11747"
    },
    {
      "label": "RAGTruth",
      "url": "https://arxiv.org/abs/2401.00396"
    },
    {
      "label": "Hallucinated references",
      "url": "https://arxiv.org/abs/2305.18248"
    },
    {
      "label": "SimpleQA",
      "url": "https://arxiv.org/abs/2411.04368"
    },
    {
      "label": "The honesty leaderboard",
      "url": "https://www.honesty.tools/leaderboard/"
    },
    {
      "label": "the surface audit",
      "url": "https://www.honesty.tools/audit"
    },
    {
      "label": "Photon, a grounding research preview",
      "url": "https://www.honesty.tools/verify"
    },
    {
      "label": "lying",
      "url": "https://www.honesty.tools/field#w-deception"
    },
    {
      "label": "MASK",
      "url": "https://arxiv.org/abs/2503.03750"
    },
    {
      "label": "Apollo Research's deception-probe datasets",
      "url": "https://arxiv.org/abs/2502.03407"
    },
    {
      "label": "probes",
      "url": "https://www.honesty.tools/field#w-probe"
    },
    {
      "label": "Truth is Universal",
      "url": "https://arxiv.org/abs/2407.12831"
    },
    {
      "label": "audit dossier",
      "url": "https://github.com/techno-optimist/honestytools/blob/main/AUDIT_DOSSIER.md"
    },
    {
      "label": "calibration",
      "url": "https://www.honesty.tools/field#w-calibration"
    },
    {
      "label": "Polish entity familiarity",
      "url": "https://arxiv.org/abs/2607.13568"
    },
    {
      "label": "Calibration preview",
      "url": "https://www.honesty.tools/calibrate"
    },
    {
      "label": "Evidence and corrections",
      "url": "https://www.honesty.tools/evidence"
    },
    {
      "label": "interpretability",
      "url": "https://www.honesty.tools/field#w-interpretability"
    },
    {
      "label": "The Geometry of Truth",
      "url": "https://arxiv.org/abs/2310.06824"
    },
    {
      "label": "Discovering Latent Knowledge",
      "url": "https://arxiv.org/abs/2212.03827"
    },
    {
      "label": "TransformerLens",
      "url": "https://github.com/TransformerLensOrg/TransformerLens"
    },
    {
      "label": "Neuronpedia",
      "url": "https://www.neuronpedia.org"
    },
    {
      "label": "Transformer Circuits",
      "url": "https://transformer-circuits.pub"
    },
    {
      "label": "surface match",
      "url": "https://www.honesty.tools/field#instruments"
    },
    {
      "label": "Geometric interpretability",
      "url": "https://www.honesty.tools/geometry"
    },
    {
      "label": "The Atlas",
      "url": "https://www.honesty.tools/atlas"
    },
    {
      "label": "Research reading room",
      "url": "https://www.honesty.tools/research"
    },
    {
      "label": "evaluation awareness",
      "url": "https://www.honesty.tools/field#w-eval-awareness"
    },
    {
      "label": "sandbagging",
      "url": "https://www.honesty.tools/field#w-sandbagging"
    },
    {
      "label": "Large Language Models Often Know When They Are Being Evaluated",
      "url": "https://arxiv.org/abs/2505.23836"
    },
    {
      "label": "AI Sandbagging",
      "url": "https://arxiv.org/abs/2406.07358"
    },
    {
      "label": "Alignment Faking in Large Language Models",
      "url": "https://arxiv.org/abs/2412.14093"
    },
    {
      "label": "sycophancy",
      "url": "https://www.honesty.tools/field#w-sycophancy"
    },
    {
      "label": "Towards Understanding Sycophancy in Language Models",
      "url": "https://arxiv.org/abs/2310.13548"
    },
    {
      "label": "the audit page",
      "url": "https://www.honesty.tools/audit"
    },
    {
      "label": "The surface audit",
      "url": "https://www.honesty.tools/audit"
    },
    {
      "label": "Evidence workbench",
      "url": "https://www.honesty.tools/evidence-methods.html"
    },
    {
      "label": "Corrections",
      "url": "https://www.honesty.tools/evidence"
    },
    {
      "label": "NullTest",
      "url": "https://github.com/techno-optimist/honestytools"
    },
    {
      "label": "nulltest_audit_results.json",
      "url": "https://www.honesty.tools/verifier/nulltest_audit_results.json"
    },
    {
      "label": "surface_baseline_results.json",
      "url": "https://www.honesty.tools/verifier/surface_baseline_results.json"
    },
    {
      "label": "honestytools repository",
      "url": "https://github.com/techno-optimist/honestytools"
    },
    {
      "label": "Download nulltest.py",
      "url": "https://www.honesty.tools/verifier/nulltest.py"
    },
    {
      "label": "Read the method",
      "url": "https://github.com/techno-optimist/honestytools#the-number-to-beat"
    },
    {
      "label": "All instruments",
      "url": "https://www.honesty.tools/tools"
    },
    {
      "label": "Apollo Research",
      "url": "https://www.apolloresearch.ai"
    },
    {
      "label": "METR",
      "url": "https://metr.org"
    },
    {
      "label": "UK AI Security Institute",
      "url": "https://www.aisi.gov.uk"
    },
    {
      "label": "Inspect",
      "url": "https://inspect.aisi.org.uk"
    },
    {
      "label": "AI Alignment Forum",
      "url": "https://www.alignmentforum.org"
    },
    {
      "label": "AISafety.info",
      "url": "https://aisafety.info"
    },
    {
      "label": "BlueDot Impact",
      "url": "https://bluedot.org"
    }
  ],
  "claims": null,
  "limitations": [
    "Static authored content from the public HTML. Interactive outputs are not simulated. Source-specific research, product and evidence limits remain in the text; this export is not new verification."
  ],
  "corrections": [],
  "supersedes": [],
  "representations": {
    "html": "https://www.honesty.tools/field",
    "markdown": "/page-text/field.md",
    "metadata": "/page-data/field.json"
  }
}