{
  "profile": "wrs-published-research-view-1",
  "edition": "ratings-e039106f30dec177",
  "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS",
  "entry": {
    "id": "weber-et-al-redpajama-transparent-data-2024",
    "title": "RedPajama: reproducible corpus mixtures and web quality signals",
    "summary": "Documented a best-effort open reproduction of the LLaMA1 data mixture, an unfiltered multilingual web pool with quality metadata and small-model ablations for constructing filtered subsets.",
    "boundary": "The 2024 technical report documents resources released earlier. Best-effort reproduction does not provide access to the original private LLaMA corpus. The pool of more than 100 trillion tokens differs from the 50-trillion-token subset carrying quality signals. No universally best filter or blanket rights clearance is established.",
    "area": "Language-model data curation",
    "topic": "Artificial intelligence",
    "year": 2024,
    "lane": "HUMAN",
    "result": "ALGORITHM",
    "package_id": "weber-et-al-redpajama-transparent-data-2024",
    "target": {
      "record_id": "urn:uuid:b68d5bf0-7230-42f5-a23d-603f899d7023",
      "version": 1,
      "content_hash": "sha256:e4e53d330f9d1f28b4366b7d84e73eaf957fa80f9d882500670c802e112b2367"
    },
    "humans": [
      "Maurice Weber",
      "Daniel Y. Fu",
      "Quentin Anthony",
      "Yonatan Oren",
      "Shane Adams",
      "Anton Alexandrov",
      "Xiaozhong Lyu",
      "Huu Nguyen",
      "Xiaozhe Yao",
      "Virginia Adams",
      "Ben Athiwaratkun",
      "Rahul Chalamala",
      "Kezhen Chen",
      "Max Ryabinin",
      "Tri Dao",
      "Percy Liang",
      "Christopher Ré",
      "Irina Rish",
      "Ce Zhang"
    ],
    "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-2411-12372v1",
        "title": "RedPajama: an Open Dataset for Training Large Language Models",
        "url": "https://arxiv.org/abs/2411.12372v1"
      }
    ],
    "points": "65000",
    "published_grade": {
      "id": "weber-et-al-redpajama-transparent-data-2024",
      "entry_id": "weber-et-al-redpajama-transparent-data-2024",
      "band": "B",
      "modifier": "+",
      "target": {
        "record_id": "urn:uuid:b68d5bf0-7230-42f5-a23d-603f899d7023",
        "version": 1,
        "content_hash": "sha256:e4e53d330f9d1f28b4366b7d84e73eaf957fa80f9d882500670c802e112b2367"
      },
      "what_changed": "Documented a best-effort open reproduction of the LLaMA1 data mixture, an unfiltered multilingual web pool with quality metadata and small-model ablations for constructing filtered subsets.",
      "scope": "Documented a best-effort open reproduction of the LLaMA1 data mixture, an unfiltered multilingual web pool with quality metadata and small-model ablations for constructing filtered subsets. The 2024 technical report documents resources released earlier. Best-effort reproduction does not provide access to the original private LLaMA corpus. The pool of more than 100 trillion tokens differs from the 50-trillion-token subset carrying quality signals. No universally best filter or blanket rights clearance is established.",
      "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:b68d5bf0-7230-42f5-a23d-603f899d7023",
        "version": 1,
        "content_hash": "sha256:e4e53d330f9d1f28b4366b7d84e73eaf957fa80f9d882500670c802e112b2367"
      },
      "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": "Maurice Weber",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Daniel Y. Fu",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Quentin Anthony",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Yonatan Oren",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Shane Adams",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Anton Alexandrov",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Xiaozhong Lyu",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Huu Nguyen",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Xiaozhe Yao",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Virginia Adams",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Ben Athiwaratkun",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Rahul Chalamala",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Kezhen Chen",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Max Ryabinin",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Tri Dao",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Percy Liang",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Christopher Ré",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Irina Rish",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Ce Zhang",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        }
      ],
      "evidence": [
        {
          "description": "Bibliographic/stated-result source; not a WRS proof verification",
          "id": "source-0",
          "kind": "source_reference",
          "source": {
            "citation": "RedPajama: an Open Dataset for Training Large Language Models",
            "url": "https://arxiv.org/abs/2411.12372v1"
          }
        }
      ],
      "origin": "HUMAN_LEGACY",
      "record_profile": "wrs-local-research-notes-0.1",
      "record_type": "research_note",
      "rrp_version": "0.1",
      "scope": {
        "description": "The 2024 technical report documents resources released earlier. Best-effort reproduction does not provide access to the original private LLaMA corpus. The pool of more than 100 trillion tokens differs from the 50-trillion-token subset carrying quality signals. No universally best filter or blanket rights clearance is established.",
        "domain": "Language-model data curation"
      },
      "sources": [
        {
          "citation": "RedPajama: an Open Dataset for Training Large Language Models",
          "url": "https://arxiv.org/abs/2411.12372v1"
        }
      ],
      "statement": "# RedPajama: reproducible corpus mixtures and web quality signals\n\nDocumented a best-effort open reproduction of the LLaMA1 data mixture, an unfiltered multilingual web pool with quality metadata and small-model ablations for constructing filtered subsets.\n\n## Scientific scope\n\nThe 2024 technical report documents resources released earlier. Best-effort reproduction does not provide access to the original private LLaMA corpus. The pool of more than 100 trillion tokens differs from the 50-trillion-token subset carrying quality signals. No universally best filter or blanket rights clearance is established.\n\n## People and systems\n\nMaurice Weber; Daniel Y. Fu; Quentin Anthony; Yonatan Oren; Shane Adams; Anton Alexandrov; Xiaozhong Lyu; Huu Nguyen; Xiaozhe Yao; Virginia Adams; Ben Athiwaratkun; Rahul Chalamala; Kezhen Chen; Max Ryabinin; Tri Dao; Percy Liang; Christopher Ré; Irina Rish; Ce Zhang\n\n\n\n## Original sources\n\n- [RedPajama: an Open Dataset for Training Large Language Models](https://arxiv.org/abs/2411.12372v1)\n",
      "summary": "Documented a best-effort open reproduction of the LLaMA1 data mixture, an unfiltered multilingual web pool with quality metadata and small-model ablations for constructing filtered subsets.",
      "title": "RedPajama: reproducible corpus mixtures and web quality signals"
    },
    "record_id": "urn:uuid:b68d5bf0-7230-42f5-a23d-603f899d7023",
    "version": 1,
    "source_ref": {
      "record_id": "urn:uuid:b68d5bf0-7230-42f5-a23d-603f899d7023",
      "version": 1,
      "content_hash": "sha256:e4e53d330f9d1f28b4366b7d84e73eaf957fa80f9d882500670c802e112b2367"
    },
    "export_view": "wrs-public-editorial-view-1",
    "public_view_hash": "sha256:76886aa477a7173fee5ddffbe7462b42b1fa8f35f8a1e20dbb567fe5cfdc278d"
  },
  "target": {
    "record_id": "urn:uuid:b68d5bf0-7230-42f5-a23d-603f899d7023",
    "version": 1,
    "content_hash": "sha256:e4e53d330f9d1f28b4366b7d84e73eaf957fa80f9d882500670c802e112b2367"
  },
  "published_grade": {
    "profile": "wrs-published-grade-1",
    "edition": "2026-10-28.1",
    "id": "weber-et-al-redpajama-transparent-data-2024",
    "entry_id": "weber-et-al-redpajama-transparent-data-2024",
    "revision": 1,
    "title": "RedPajama: reproducible corpus mixtures and web quality signals",
    "context": "Language-model data curation",
    "track": "Human-designed research method or resource",
    "band": "B",
    "modifier": "+",
    "status": "PROVISIONAL_EDITORIAL",
    "assessed_on": "2026-10-16",
    "evidence_cutoff": "2026-10-16",
    "achievement_state": "SUPPORTED_SCOPED_RESULT",
    "what_changed": "Documented a best-effort open reproduction of the LLaMA1 data mixture, an unfiltered multilingual web pool with quality metadata and small-model ablations for constructing filtered subsets.",
    "scope": "Documented a best-effort open reproduction of the LLaMA1 data mixture, an unfiltered multilingual web pool with quality metadata and small-model ablations for constructing filtered subsets. The 2024 technical report documents resources released earlier. Best-effort reproduction does not provide access to the original private LLaMA corpus. The pool of more than 100 trillion tokens differs from the 50-trillion-token subset carrying quality signals. No universally best filter or blanket rights clearance is established.",
    "evidence_summary": "The linked original sources describe this contribution and its evidence limits. WRS has not independently reconstructed the full scientific claims. The 2024 technical report documents resources released earlier. Best-effort reproduction does not provide access to the original private LLaMA corpus. The pool of more than 100 trillion tokens differs from the 50-trillion-token subset carrying quality signals. No universally best filter or blanket rights clearance is established.",
    "references": [
      {
        "title": "RedPajama: an Open Dataset for Training Large Language Models",
        "url": "https://arxiv.org/abs/2411.12372v1",
        "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:b68d5bf0-7230-42f5-a23d-603f899d7023",
      "version": 1,
      "content_hash": "sha256:e4e53d330f9d1f28b4366b7d84e73eaf957fa80f9d882500670c802e112b2367"
    },
    "scientific_correctness_certified": false,
    "independent_expert_calibration": false,
    "method": "PROPRIETARY",
    "points": "65000"
  },
  "canonical_agent_submission": false,
  "scientific_verification_awarded": false,
  "reader_guide": null,
  "reading_status": "SOURCE_LINKED_NOTE"
}