{
  "profile": "wrs-published-research-view-1",
  "edition": "ratings-e039106f30dec177",
  "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS",
  "entry": {
    "id": "zhou-et-al-transfusion-2024",
    "title": "Transfusion: mixed discrete-token prediction and continuous-image diffusion",
    "summary": "Trained one transformer on interleaved text and continuous image latents with next-token text loss, image diffusion loss and causal-text/bidirectional-image attention.",
    "boundary": "Not first multimodal modeling or diffusion. Acquired delta is the joint recipe/controlled integration, not a claim of information-lossless pixels: image encoding and patch compression remain. Controlled Chameleon-style comparisons are not blanket superiority over every trained model.",
    "area": "Unified multimodal generative modeling",
    "topic": "Artificial intelligence",
    "year": 2024,
    "lane": "HUMAN",
    "result": "ALGORITHM",
    "package_id": "zhou-et-al-transfusion-2024",
    "target": {
      "record_id": "urn:uuid:6e9a14e7-7cb4-4eb8-a3f8-4e5c62c321e3",
      "version": 1,
      "content_hash": "sha256:68fb84f2a3c0d08329e3ea4864a53ca505813cf17689a4038d5e306e96e42182"
    },
    "humans": [
      "Chunting Zhou",
      "Lili Yu",
      "Arun Babu",
      "Kushal Tirumala",
      "Michihiro Yasunaga",
      "Leonid Shamis",
      "Jacob Kahn",
      "Xuezhe Ma",
      "Luke Zettlemoyer",
      "Omer Levy"
    ],
    "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-2408-11039v1",
        "title": "Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model",
        "url": "https://arxiv.org/abs/2408.11039v1"
      }
    ],
    "points": "125000",
    "published_grade": {
      "id": "zhou-et-al-transfusion-2024",
      "entry_id": "zhou-et-al-transfusion-2024",
      "band": "A",
      "modifier": null,
      "target": {
        "record_id": "urn:uuid:6e9a14e7-7cb4-4eb8-a3f8-4e5c62c321e3",
        "version": 1,
        "content_hash": "sha256:68fb84f2a3c0d08329e3ea4864a53ca505813cf17689a4038d5e306e96e42182"
      },
      "what_changed": "Trained one transformer on interleaved text and continuous image latents with next-token text loss, image diffusion loss and causal-text/bidirectional-image attention.",
      "scope": "Trained one transformer on interleaved text and continuous image latents with next-token text loss, image diffusion loss and causal-text/bidirectional-image attention. Not first multimodal modeling or diffusion. Acquired delta is the joint recipe/controlled integration, not a claim of information-lossless pixels: image encoding and patch compression remain. Controlled Chameleon-style comparisons are not blanket superiority over every trained model.",
      "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:6e9a14e7-7cb4-4eb8-a3f8-4e5c62c321e3",
        "version": 1,
        "content_hash": "sha256:68fb84f2a3c0d08329e3ea4864a53ca505813cf17689a4038d5e306e96e42182"
      },
      "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": "Chunting Zhou",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model",
            "url": "https://arxiv.org/abs/2408.11039v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Lili Yu",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model",
            "url": "https://arxiv.org/abs/2408.11039v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Arun Babu",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model",
            "url": "https://arxiv.org/abs/2408.11039v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Kushal Tirumala",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model",
            "url": "https://arxiv.org/abs/2408.11039v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Michihiro Yasunaga",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model",
            "url": "https://arxiv.org/abs/2408.11039v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Leonid Shamis",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model",
            "url": "https://arxiv.org/abs/2408.11039v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Jacob Kahn",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model",
            "url": "https://arxiv.org/abs/2408.11039v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Xuezhe Ma",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model",
            "url": "https://arxiv.org/abs/2408.11039v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Luke Zettlemoyer",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model",
            "url": "https://arxiv.org/abs/2408.11039v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Omer Levy",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model",
            "url": "https://arxiv.org/abs/2408.11039v1"
          }
        }
      ],
      "evidence": [
        {
          "description": "Bibliographic/stated-result source; not a WRS proof verification",
          "id": "source-0",
          "kind": "source_reference",
          "source": {
            "citation": "Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model",
            "url": "https://arxiv.org/abs/2408.11039v1"
          }
        }
      ],
      "origin": "HUMAN_LEGACY",
      "record_profile": "wrs-local-research-notes-0.1",
      "record_type": "research_note",
      "rrp_version": "0.1",
      "scope": {
        "description": "Not first multimodal modeling or diffusion. Acquired delta is the joint recipe/controlled integration, not a claim of information-lossless pixels: image encoding and patch compression remain. Controlled Chameleon-style comparisons are not blanket superiority over every trained model.",
        "domain": "Unified multimodal generative modeling"
      },
      "sources": [
        {
          "citation": "Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model",
          "url": "https://arxiv.org/abs/2408.11039v1"
        }
      ],
      "statement": "# Transfusion: mixed discrete-token prediction and continuous-image diffusion\n\nTrained one transformer on interleaved text and continuous image latents with next-token text loss, image diffusion loss and causal-text/bidirectional-image attention.\n\n## Scientific scope\n\nNot first multimodal modeling or diffusion. Acquired delta is the joint recipe/controlled integration, not a claim of information-lossless pixels: image encoding and patch compression remain. Controlled Chameleon-style comparisons are not blanket superiority over every trained model.\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\nChunting Zhou; Lili Yu; Arun Babu; Kushal Tirumala; Michihiro Yasunaga; Leonid Shamis; Jacob Kahn; Xuezhe Ma; Luke Zettlemoyer; Omer Levy\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- [Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model](https://arxiv.org/abs/2408.11039v1)\n",
      "summary": "Trained one transformer on interleaved text and continuous image latents with next-token text loss, image diffusion loss and causal-text/bidirectional-image attention.",
      "title": "Transfusion: mixed discrete-token prediction and continuous-image diffusion"
    },
    "record_id": "urn:uuid:6e9a14e7-7cb4-4eb8-a3f8-4e5c62c321e3",
    "version": 1,
    "source_ref": {
      "record_id": "urn:uuid:6e9a14e7-7cb4-4eb8-a3f8-4e5c62c321e3",
      "version": 1,
      "content_hash": "sha256:68fb84f2a3c0d08329e3ea4864a53ca505813cf17689a4038d5e306e96e42182"
    },
    "export_view": "wrs-public-editorial-view-1",
    "public_view_hash": "sha256:14f76ba5c6654aa69d848f9f3246fbbdb953caf8144a9aa1963d9775f8a61e1e"
  },
  "target": {
    "record_id": "urn:uuid:6e9a14e7-7cb4-4eb8-a3f8-4e5c62c321e3",
    "version": 1,
    "content_hash": "sha256:68fb84f2a3c0d08329e3ea4864a53ca505813cf17689a4038d5e306e96e42182"
  },
  "published_grade": {
    "profile": "wrs-published-grade-1",
    "edition": "2026-10-28.1",
    "id": "zhou-et-al-transfusion-2024",
    "entry_id": "zhou-et-al-transfusion-2024",
    "revision": 1,
    "title": "Transfusion: mixed discrete-token prediction and continuous-image diffusion",
    "context": "Unified multimodal generative modeling",
    "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": "Trained one transformer on interleaved text and continuous image latents with next-token text loss, image diffusion loss and causal-text/bidirectional-image attention.",
    "scope": "Trained one transformer on interleaved text and continuous image latents with next-token text loss, image diffusion loss and causal-text/bidirectional-image attention. Not first multimodal modeling or diffusion. Acquired delta is the joint recipe/controlled integration, not a claim of information-lossless pixels: image encoding and patch compression remain. Controlled Chameleon-style comparisons are not blanket superiority over every trained model.",
    "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. Not first multimodal modeling or diffusion. Acquired delta is the joint recipe/controlled integration, not a claim of information-lossless pixels: image encoding and patch compression remain. Controlled Chameleon-style comparisons are not blanket superiority over every trained model.",
    "references": [
      {
        "title": "Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model",
        "url": "https://arxiv.org/abs/2408.11039v1",
        "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:6e9a14e7-7cb4-4eb8-a3f8-4e5c62c321e3",
      "version": 1,
      "content_hash": "sha256:68fb84f2a3c0d08329e3ea4864a53ca505813cf17689a4038d5e306e96e42182"
    },
    "scientific_correctness_certified": false,
    "independent_expert_calibration": false,
    "method": "PROPRIETARY",
    "points": "125000"
  },
  "canonical_agent_submission": false,
  "scientific_verification_awarded": false,
  "reader_guide": null,
  "reading_status": "SOURCE_LINKED_NOTE"
}