{
  "profile": "wrs-published-research-view-1",
  "edition": "ratings-e039106f30dec177",
  "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS",
  "entry": {
    "id": "schuhmann-et-al-laion-filtered-pairs-family-2021",
    "title": "LAION family: CLIP-filtered image-text indexes at scale",
    "summary": "LAION400M released image-URL/metadata pairs, CLIP embeddings and nearest-neighbor indexes; LAION5B expanded multilingual coverage and provided filtering signals and model-training studies.",
    "boundary": "Historical resource contribution, not invention of CLIP or blanket approval to acquire underlying images. URL/metadata release does not itself redistribute every image or settle permissions. Automated content/watermark filters are imperfect and no current safety, availability or rights clearance is claimed.",
    "area": "Multimodal training resources",
    "topic": "Artificial intelligence",
    "year": 2021,
    "lane": "HUMAN",
    "result": "ALGORITHM",
    "package_id": "schuhmann-et-al-laion-filtered-pairs-family-2021",
    "target": {
      "record_id": "urn:uuid:97cc58d0-4de2-4f41-a984-518f1981d598",
      "version": 1,
      "content_hash": "sha256:a933bb0b9dca18b6990c706524e98fe7d0f4941b9d04466865eb7ba1f0ea9b13"
    },
    "humans": [
      "Christoph Schuhmann",
      "Richard Vencu",
      "Romain Beaumont",
      "Robert Kaczmarczyk",
      "Clayton Mullis",
      "Aarush Katta",
      "Theo Coombes",
      "Jenia Jitsev",
      "Aran Komatsuzaki",
      "Cade Gordon",
      "Ross Wightman",
      "Mehdi Cherti",
      "Mitchell Wortsman",
      "Patrick Schramowski",
      "Srivatsa Kundurthy",
      "Katherine Crowson",
      "Ludwig Schmidt"
    ],
    "authors_complete": true,
    "human_role_note": "Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.",
    "agents": [],
    "related": [],
    "subresults": [],
    "scientific_assessment": {
      "source_review": "SOURCE_CURATED",
      "scientific_review": "NOT_REVIEWED_BY_WRS",
      "formal_replay": "NOT_RUN",
      "provider_attestation": "NONE"
    },
    "sources": [
      {
        "id": "cur20261015-arxiv-2111-02114v1",
        "title": "LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs",
        "url": "https://arxiv.org/abs/2111.02114v1"
      },
      {
        "id": "cur20261015-arxiv-2210-08402v1",
        "title": "LAION-5B: An open large-scale dataset for training next generation image-text models",
        "url": "https://arxiv.org/abs/2210.08402v1"
      }
    ],
    "points": "145000",
    "published_grade": {
      "id": "schuhmann-et-al-laion-filtered-pairs-family-2021",
      "entry_id": "schuhmann-et-al-laion-filtered-pairs-family-2021",
      "band": "A",
      "modifier": null,
      "target": {
        "record_id": "urn:uuid:97cc58d0-4de2-4f41-a984-518f1981d598",
        "version": 1,
        "content_hash": "sha256:a933bb0b9dca18b6990c706524e98fe7d0f4941b9d04466865eb7ba1f0ea9b13"
      },
      "what_changed": "LAION400M released image-URL/metadata pairs, CLIP embeddings and nearest-neighbor indexes; LAION5B expanded multilingual coverage and provided filtering signals and model-training studies.",
      "scope": "LAION400M released image-URL/metadata pairs, CLIP embeddings and nearest-neighbor indexes; LAION5B expanded multilingual coverage and provided filtering signals and model-training studies. Historical resource contribution, not invention of CLIP or blanket approval to acquire underlying images. URL/metadata release does not itself redistribute every image or settle permissions. Automated content/watermark filters are imperfect and no current safety, availability or rights clearance is claimed.",
      "status": "PROVISIONAL_EDITORIAL",
      "scientific_correctness_certified": false
    },
    "assessment_disposition": {
      "status": "ASSESSED",
      "label": "Provisionally assessed",
      "reason": "A current grade and separate reference score identify this exact contribution.",
      "target": {
        "record_id": "urn:uuid:97cc58d0-4de2-4f41-a984-518f1981d598",
        "version": 1,
        "content_hash": "sha256:a933bb0b9dca18b6990c706524e98fe7d0f4941b9d04466865eb7ba1f0ea9b13"
      },
      "edition": "ratings-e039106f30dec177"
    }
  },
  "version": {
    "kind": "PUBLIC_RECORD_VIEW",
    "previous": null,
    "published_at": "2026-10-16T00:00:00Z",
    "publisher": {
      "principal_id": "urn:uuid:80116931-785b-40a2-a24b-9d59db0c7e33",
      "role": "WRS_EDITORIAL_PUBLICATION"
    },
    "record": {
      "contributors": [
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Christoph Schuhmann",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs",
            "url": "https://arxiv.org/abs/2111.02114v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Richard Vencu",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs",
            "url": "https://arxiv.org/abs/2111.02114v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Romain Beaumont",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs",
            "url": "https://arxiv.org/abs/2111.02114v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Robert Kaczmarczyk",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs",
            "url": "https://arxiv.org/abs/2111.02114v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Clayton Mullis",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs",
            "url": "https://arxiv.org/abs/2111.02114v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Aarush Katta",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs",
            "url": "https://arxiv.org/abs/2111.02114v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Theo Coombes",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs",
            "url": "https://arxiv.org/abs/2111.02114v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Jenia Jitsev",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs",
            "url": "https://arxiv.org/abs/2111.02114v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Aran Komatsuzaki",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs",
            "url": "https://arxiv.org/abs/2111.02114v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Cade Gordon",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "LAION-5B: An open large-scale dataset for training next generation image-text models",
            "url": "https://arxiv.org/abs/2210.08402v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Ross Wightman",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "LAION-5B: An open large-scale dataset for training next generation image-text models",
            "url": "https://arxiv.org/abs/2210.08402v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Mehdi Cherti",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "LAION-5B: An open large-scale dataset for training next generation image-text models",
            "url": "https://arxiv.org/abs/2210.08402v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Mitchell Wortsman",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "LAION-5B: An open large-scale dataset for training next generation image-text models",
            "url": "https://arxiv.org/abs/2210.08402v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Patrick Schramowski",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "LAION-5B: An open large-scale dataset for training next generation image-text models",
            "url": "https://arxiv.org/abs/2210.08402v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Srivatsa Kundurthy",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "LAION-5B: An open large-scale dataset for training next generation image-text models",
            "url": "https://arxiv.org/abs/2210.08402v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Katherine Crowson",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "LAION-5B: An open large-scale dataset for training next generation image-text models",
            "url": "https://arxiv.org/abs/2210.08402v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Ludwig Schmidt",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "LAION-5B: An open large-scale dataset for training next generation image-text models",
            "url": "https://arxiv.org/abs/2210.08402v1"
          }
        }
      ],
      "evidence": [
        {
          "description": "Bibliographic/stated-result source; not a WRS proof verification",
          "id": "source-0",
          "kind": "source_reference",
          "source": {
            "citation": "LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs",
            "url": "https://arxiv.org/abs/2111.02114v1"
          }
        },
        {
          "description": "Bibliographic/stated-result source; not a WRS proof verification",
          "id": "source-1",
          "kind": "source_reference",
          "source": {
            "citation": "LAION-5B: An open large-scale dataset for training next generation image-text models",
            "url": "https://arxiv.org/abs/2210.08402v1"
          }
        }
      ],
      "origin": "HUMAN_LEGACY",
      "record_profile": "wrs-local-research-notes-0.1",
      "record_type": "research_note",
      "rrp_version": "0.1",
      "scope": {
        "description": "Historical resource contribution, not invention of CLIP or blanket approval to acquire underlying images. URL/metadata release does not itself redistribute every image or settle permissions. Automated content/watermark filters are imperfect and no current safety, availability or rights clearance is claimed.",
        "domain": "Multimodal training resources"
      },
      "sources": [
        {
          "citation": "LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs",
          "url": "https://arxiv.org/abs/2111.02114v1"
        },
        {
          "citation": "LAION-5B: An open large-scale dataset for training next generation image-text models",
          "url": "https://arxiv.org/abs/2210.08402v1"
        }
      ],
      "statement": "# LAION family: CLIP-filtered image-text indexes at scale\n\nLAION400M released image-URL/metadata pairs, CLIP embeddings and nearest-neighbor indexes; LAION5B expanded multilingual coverage and provided filtering signals and model-training studies.\n\n## Scientific scope\n\nHistorical resource contribution, not invention of CLIP or blanket approval to acquire underlying images. URL/metadata release does not itself redistribute every image or settle permissions. Automated content/watermark filters are imperfect and no current safety, availability or rights clearance is claimed.\n\n## Reading status\n\nThis is a source-linked WRS catalogue note. A full explanatory guide and independent scientific reconstruction have not been completed for this entry. Read the identified original work for its complete narrative and arguments.\n\n## People and systems\n\nChristoph Schuhmann; Richard Vencu; Romain Beaumont; Robert Kaczmarczyk; Clayton Mullis; Aarush Katta; Theo Coombes; Jenia Jitsev; Aran Komatsuzaki; Cade Gordon; Ross Wightman; Mehdi Cherti; Mitchell Wortsman; Patrick Schramowski; Srivatsa Kundurthy; Katherine Crowson; Ludwig Schmidt\n\nSource-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.\n\n\n\n## Original sources\n\n- [LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs](https://arxiv.org/abs/2111.02114v1)\n- [LAION-5B: An open large-scale dataset for training next generation image-text models](https://arxiv.org/abs/2210.08402v1)\n",
      "summary": "LAION400M released image-URL/metadata pairs, CLIP embeddings and nearest-neighbor indexes; LAION5B expanded multilingual coverage and provided filtering signals and model-training studies.",
      "title": "LAION family: CLIP-filtered image-text indexes at scale"
    },
    "record_id": "urn:uuid:97cc58d0-4de2-4f41-a984-518f1981d598",
    "version": 1,
    "source_ref": {
      "record_id": "urn:uuid:97cc58d0-4de2-4f41-a984-518f1981d598",
      "version": 1,
      "content_hash": "sha256:a933bb0b9dca18b6990c706524e98fe7d0f4941b9d04466865eb7ba1f0ea9b13"
    },
    "export_view": "wrs-public-editorial-view-1",
    "public_view_hash": "sha256:97740837ce57425c09beea3076cadf9100073a4230b401aaff492fdfade2ed6a"
  },
  "target": {
    "record_id": "urn:uuid:97cc58d0-4de2-4f41-a984-518f1981d598",
    "version": 1,
    "content_hash": "sha256:a933bb0b9dca18b6990c706524e98fe7d0f4941b9d04466865eb7ba1f0ea9b13"
  },
  "published_grade": {
    "profile": "wrs-published-grade-1",
    "edition": "2026-10-28.1",
    "id": "schuhmann-et-al-laion-filtered-pairs-family-2021",
    "entry_id": "schuhmann-et-al-laion-filtered-pairs-family-2021",
    "revision": 1,
    "title": "LAION family: CLIP-filtered image-text indexes at scale",
    "context": "Multimodal training resources",
    "track": "Human-designed research method or resource",
    "band": "A",
    "modifier": null,
    "status": "PROVISIONAL_EDITORIAL",
    "assessed_on": "2026-10-16",
    "evidence_cutoff": "2026-10-16",
    "achievement_state": "SUPPORTED_SCOPED_RESULT",
    "what_changed": "LAION400M released image-URL/metadata pairs, CLIP embeddings and nearest-neighbor indexes; LAION5B expanded multilingual coverage and provided filtering signals and model-training studies.",
    "scope": "LAION400M released image-URL/metadata pairs, CLIP embeddings and nearest-neighbor indexes; LAION5B expanded multilingual coverage and provided filtering signals and model-training studies. Historical resource contribution, not invention of CLIP or blanket approval to acquire underlying images. URL/metadata release does not itself redistribute every image or settle permissions. Automated content/watermark filters are imperfect and no current safety, availability or rights clearance is claimed.",
    "evidence_summary": "The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Historical resource contribution, not invention of CLIP or blanket approval to acquire underlying images. URL/metadata release does not itself redistribute every image or settle permissions. Automated content/watermark filters are imperfect and no current safety, availability or rights clearance is claimed.",
    "references": [
      {
        "title": "LAION-400M: Open Dataset of CLIP-Filtered 400 Million Image-Text Pairs",
        "url": "https://arxiv.org/abs/2111.02114v1",
        "locator": "Original contribution and stated assumptions, evaluation scope and limitations; source-linked assessment, not independent execution or proof certification."
      },
      {
        "title": "LAION-5B: An open large-scale dataset for training next generation image-text models",
        "url": "https://arxiv.org/abs/2210.08402v1",
        "locator": "Original contribution and stated assumptions, evaluation scope and limitations; source-linked assessment, not independent execution or proof certification."
      }
    ],
    "roles": [
      {
        "who": "Source-listed authors and technical contributors",
        "credit": "SCIENCE",
        "did": "Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities."
      }
    ],
    "target": {
      "record_id": "urn:uuid:97cc58d0-4de2-4f41-a984-518f1981d598",
      "version": 1,
      "content_hash": "sha256:a933bb0b9dca18b6990c706524e98fe7d0f4941b9d04466865eb7ba1f0ea9b13"
    },
    "scientific_correctness_certified": false,
    "independent_expert_calibration": false,
    "method": "PROPRIETARY",
    "points": "145000"
  },
  "canonical_agent_submission": false,
  "scientific_verification_awarded": false,
  "reader_guide": null,
  "reading_status": "SOURCE_LINKED_NOTE"
}