Redwood Research

{
  "id": "redwood-research",
  "wikiId": "E557",
  "path": "/knowledge-base/organizations/redwood-research/",
  "filePath": "knowledge-base/organizations/redwood-research.mdx",
  "title": "Redwood Research",
  "quality": 78,
  "readerImportance": 31.5,
  "researchImportance": 45,
  "tacticalValue": null,
  "contentFormat": "article",
  "causalLevel": null,
  "lastUpdated": "2026-01-31",
  "dateCreated": "2026-02-15",
  "summary": null,
  "description": "A nonprofit AI safety and security research organization founded in 2021, known for pioneering AI Control research, developing causal scrubbing interpretability methods, and conducting landmark alignment faking studies with Anthropic.",
  "ratings": {
    "novelty": 7,
    "rigor": 7,
    "completeness": 7,
    "actionability": 5
  },
  "category": "organizations",
  "subcategory": "safety-orgs",
  "clusters": [
    "ai-safety",
    "community"
  ],
  "metrics": {
    "wordCount": 1397,
    "tableCount": 7,
    "diagramCount": 0,
    "internalLinks": 25,
    "externalLinks": 15,
    "footnoteCount": 32,
    "bulletRatio": 0,
    "sectionCount": 24,
    "hasOverview": true,
    "structuralScore": 14
  },
  "suggestedQuality": 93,
  "updateFrequency": 21,
  "evergreen": true,
  "wordCount": 1397,
  "unconvertedLinks": [
    {
      "text": "EA Forum",
      "url": "https://forum.effectivealtruism.org/posts/MGbdhjgd2v6cg3vjv/apply-to-the-redwood-research-mechanistic-interpretability",
      "resourceId": "58b17dcb71ac7212",
      "resourceTitle": "Apply to the Redwood Research Mechanistic Interpretability Experiment (REMIX), a research program in Berkeley"
    },
    {
      "text": "arXiv",
      "url": "https://arxiv.org/pdf/2312.06942",
      "resourceId": "cc80ab28579c5794",
      "resourceTitle": "Redwood Research's AI Control paper (December 2023)"
    },
    {
      "text": "Anthropic",
      "url": "https://www.anthropic.com/research/alignment-faking",
      "resourceId": "c2cfd72baafd64a9",
      "resourceTitle": "Anthropic's 2024 alignment faking study"
    },
    {
      "text": "arXiv",
      "url": "https://arxiv.org/abs/2405.19550",
      "resourceId": "2bb15dbd965ba994",
      "resourceTitle": "Stress-Testing Capability Elicitation With Password-Locked Models - arXiv"
    },
    {
      "text": "arXiv",
      "url": "https://arxiv.org/abs/2501.17315",
      "resourceId": "56bae0415b228464",
      "resourceTitle": "A Sketch of an AI Control Safety Case - arXiv"
    },
    {
      "text": "arXiv",
      "url": "https://arxiv.org/abs/2504.10374",
      "resourceId": "1cbfd6cb09ce2b50",
      "resourceTitle": "Ctrl-Z: Controlling AI Agents via Resampling - arXiv"
    },
    {
      "text": "arXiv",
      "url": "https://arxiv.org/abs/2512.15688",
      "resourceId": "a4efd028ec6d9614",
      "resourceTitle": "BashArena: A Control Setting for Highly Privileged AI Agents - arXiv"
    },
    {
      "text": "OP Grants",
      "url": "https://www.openphilanthropy.org/grants/redwood-research-general-support/",
      "resourceId": "8c79e00bab007a63",
      "resourceTitle": "over $9.4 million from Open Philanthropy"
    },
    {
      "text": "OP Grants",
      "url": "https://www.openphilanthropy.org/grants/",
      "resourceId": "2fcdf851ed57384c",
      "resourceTitle": "Open Philanthropy Grants Database"
    },
    {
      "text": "ProPublica 990",
      "url": "https://projects.propublica.org/nonprofits/organizations/871702255",
      "resourceId": "kb-852dca560bf056cb",
      "resourceTitle": "Redwood Research Group Inc - Nonprofit Explorer - ProPublica"
    },
    {
      "text": "ProPublica 990",
      "url": "https://projects.propublica.org/nonprofits/organizations/871702255",
      "resourceId": "kb-852dca560bf056cb",
      "resourceTitle": "Redwood Research Group Inc - Nonprofit Explorer - ProPublica"
    },
    {
      "text": "ProPublica 990",
      "url": "https://projects.propublica.org/nonprofits/organizations/871702255",
      "resourceId": "kb-852dca560bf056cb",
      "resourceTitle": "Redwood Research Group Inc - Nonprofit Explorer - ProPublica"
    }
  ],
  "unconvertedLinkCount": 12,
  "convertedLinkCount": 0,
  "backlinkCount": 55,
  "citationHealth": {
    "total": 45,
    "withQuotes": 30,
    "verified": 30,
    "accuracyChecked": 30,
    "accurate": 18,
    "inaccurate": 0,
    "avgScore": 0.9502938191095988
  },
  "hallucinationRisk": {
    "level": "low",
    "score": 30,
    "factors": [
      "biographical-claims",
      "well-cited",
      "high-rigor"
    ]
  },
  "entityType": "organization",
  "redundancy": {
    "maxSimilarity": 13,
    "similarPages": [
      {
        "id": "ai-control",
        "title": "AI Control",
        "path": "/knowledge-base/responses/ai-control/",
        "similarity": 13
      },
      {
        "id": "apollo-research",
        "title": "Apollo Research",
        "path": "/knowledge-base/organizations/apollo-research/",
        "similarity": 12
      },
      {
        "id": "secure-ai-project",
        "title": "Secure AI Project",
        "path": "/knowledge-base/organizations/secure-ai-project/",
        "similarity": 12
      },
      {
        "id": "ssi",
        "title": "Safe Superintelligence Inc. (SSI)",
        "path": "/knowledge-base/organizations/ssi/",
        "similarity": 12
      },
      {
        "id": "ilya-sutskever",
        "title": "Ilya Sutskever",
        "path": "/knowledge-base/people/ilya-sutskever/",
        "similarity": 12
      }
    ]
  },
  "coverage": {
    "passing": 8,
    "total": 13,
    "targets": {
      "tables": 6,
      "diagrams": 1,
      "internalLinks": 11,
      "externalLinks": 7,
      "footnotes": 4,
      "references": 4
    },
    "actuals": {
      "tables": 7,
      "diagrams": 0,
      "internalLinks": 25,
      "externalLinks": 15,
      "footnotes": 32,
      "references": 10,
      "quotesWithQuotes": 30,
      "quotesTotal": 45,
      "accuracyChecked": 30,
      "accuracyTotal": 45
    },
    "items": {
      "summary": "red",
      "schedule": "green",
      "entity": "green",
      "editHistory": "red",
      "overview": "green",
      "tables": "green",
      "diagrams": "red",
      "internalLinks": "green",
      "externalLinks": "green",
      "footnotes": "green",
      "references": "green",
      "quotes": "amber",
      "accuracy": "amber"
    },
    "ratingsString": "N:7 R:7 A:5 C:7"
  },
  "readerRank": 437,
  "researchRank": 302,
  "recommendedScore": 175.59
}

{
  "lesswrong": "https://www.lesswrong.com/tag/redwood-research",
  "eaForum": "https://forum.effectivealtruism.org/topics/redwood-research",
  "grokipedia": "https://grokipedia.com/page/Redwood_Research"
}