{
  "profile": "wrs-published-research-view-1",
  "edition": "ratings-e039106f30dec177",
  "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS",
  "entry": {
    "id": "chainer-define-by-run-2019-account",
    "title": "Chainer dynamic graphs, memory-aware autodiff and CuPy integration",
    "summary": "Documents define-by-run model construction, memory-aware reverse/double backpropagation and NumPy-like CuPy GPU integration in the Chainer framework.",
    "boundary": "The 2019 account retrospectively describes the 2015 Chainer launch and the 2017 separation of CuPy. Later capabilities are not backdated. Dynamic computation and automatic differentiation have predecessors; the contribution is their define-by-run framework integration.",
    "area": "Deep learning software",
    "topic": "Artificial intelligence",
    "year": 2019,
    "lane": "HUMAN",
    "result": "ALGORITHM",
    "package_id": "chainer-define-by-run-2019-account",
    "target": {
      "record_id": "urn:uuid:7f4a0b2f-9007-49b7-ad95-24ff821bf484",
      "version": 1,
      "content_hash": "sha256:54131e00af1fb617ed2983efd9a15416e4494357e2c4297251e78477f7573f91"
    },
    "humans": [
      "Seiya Tokui",
      "Ryosuke Okuta",
      "Takuya Akiba",
      "Yusuke Niitani",
      "Toru Ogawa",
      "Shunta Saito",
      "Shuji Suzuki",
      "Kota Uenishi",
      "Brian Vogel",
      "Hiroyuki Yamazaki Vincent"
    ],
    "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": "ai-oct15-arxiv-1908-00213v1",
        "title": "Chainer: A Deep Learning Framework for Accelerating the Research Cycle",
        "url": "https://arxiv.org/pdf/1908.00213v1"
      }
    ],
    "points": "140000",
    "published_grade": {
      "id": "chainer-define-by-run-2019-account",
      "entry_id": "chainer-define-by-run-2019-account",
      "band": "A",
      "modifier": null,
      "target": {
        "record_id": "urn:uuid:7f4a0b2f-9007-49b7-ad95-24ff821bf484",
        "version": 1,
        "content_hash": "sha256:54131e00af1fb617ed2983efd9a15416e4494357e2c4297251e78477f7573f91"
      },
      "what_changed": "Documents define-by-run model construction, memory-aware reverse/double backpropagation and NumPy-like CuPy GPU integration in the Chainer framework.",
      "scope": "Documents define-by-run model construction, memory-aware reverse/double backpropagation and NumPy-like CuPy GPU integration in the Chainer framework. The 2019 account retrospectively describes the 2015 Chainer launch and the 2017 separation of CuPy. Later capabilities are not backdated. Dynamic computation and automatic differentiation have predecessors; the contribution is their define-by-run framework integration.",
      "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:7f4a0b2f-9007-49b7-ad95-24ff821bf484",
        "version": 1,
        "content_hash": "sha256:54131e00af1fb617ed2983efd9a15416e4494357e2c4297251e78477f7573f91"
      },
      "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": "Seiya Tokui",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Chainer: A Deep Learning Framework for Accelerating the Research Cycle",
            "url": "https://arxiv.org/pdf/1908.00213v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Ryosuke Okuta",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Chainer: A Deep Learning Framework for Accelerating the Research Cycle",
            "url": "https://arxiv.org/pdf/1908.00213v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Takuya Akiba",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Chainer: A Deep Learning Framework for Accelerating the Research Cycle",
            "url": "https://arxiv.org/pdf/1908.00213v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Yusuke Niitani",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Chainer: A Deep Learning Framework for Accelerating the Research Cycle",
            "url": "https://arxiv.org/pdf/1908.00213v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Toru Ogawa",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Chainer: A Deep Learning Framework for Accelerating the Research Cycle",
            "url": "https://arxiv.org/pdf/1908.00213v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Shunta Saito",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Chainer: A Deep Learning Framework for Accelerating the Research Cycle",
            "url": "https://arxiv.org/pdf/1908.00213v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Shuji Suzuki",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Chainer: A Deep Learning Framework for Accelerating the Research Cycle",
            "url": "https://arxiv.org/pdf/1908.00213v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Kota Uenishi",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Chainer: A Deep Learning Framework for Accelerating the Research Cycle",
            "url": "https://arxiv.org/pdf/1908.00213v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Brian Vogel",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Chainer: A Deep Learning Framework for Accelerating the Research Cycle",
            "url": "https://arxiv.org/pdf/1908.00213v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Hiroyuki Yamazaki Vincent",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Chainer: A Deep Learning Framework for Accelerating the Research Cycle",
            "url": "https://arxiv.org/pdf/1908.00213v1"
          }
        }
      ],
      "evidence": [
        {
          "description": "Bibliographic/stated-result source; not a WRS proof verification",
          "id": "source-0",
          "kind": "source_reference",
          "source": {
            "citation": "Chainer: A Deep Learning Framework for Accelerating the Research Cycle",
            "url": "https://arxiv.org/pdf/1908.00213v1"
          }
        }
      ],
      "origin": "HUMAN_LEGACY",
      "record_profile": "wrs-local-research-notes-0.1",
      "record_type": "research_note",
      "rrp_version": "0.1",
      "scope": {
        "description": "The 2019 account retrospectively describes the 2015 Chainer launch and the 2017 separation of CuPy. Later capabilities are not backdated. Dynamic computation and automatic differentiation have predecessors; the contribution is their define-by-run framework integration.",
        "domain": "Deep learning software"
      },
      "sources": [
        {
          "citation": "Chainer: A Deep Learning Framework for Accelerating the Research Cycle",
          "url": "https://arxiv.org/pdf/1908.00213v1"
        }
      ],
      "statement": "# Chainer dynamic graphs, memory-aware autodiff and CuPy integration\n\nDocuments define-by-run model construction, memory-aware reverse/double backpropagation and NumPy-like CuPy GPU integration in the Chainer framework.\n\n## Scientific scope\n\nThe 2019 account retrospectively describes the 2015 Chainer launch and the 2017 separation of CuPy. Later capabilities are not backdated. Dynamic computation and automatic differentiation have predecessors; the contribution is their define-by-run framework integration.\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\nSeiya Tokui; Ryosuke Okuta; Takuya Akiba; Yusuke Niitani; Toru Ogawa; Shunta Saito; Shuji Suzuki; Kota Uenishi; Brian Vogel; Hiroyuki Yamazaki Vincent\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- [Chainer: A Deep Learning Framework for Accelerating the Research Cycle](https://arxiv.org/pdf/1908.00213v1)\n",
      "summary": "Documents define-by-run model construction, memory-aware reverse/double backpropagation and NumPy-like CuPy GPU integration in the Chainer framework.",
      "title": "Chainer dynamic graphs, memory-aware autodiff and CuPy integration"
    },
    "record_id": "urn:uuid:7f4a0b2f-9007-49b7-ad95-24ff821bf484",
    "version": 1,
    "source_ref": {
      "record_id": "urn:uuid:7f4a0b2f-9007-49b7-ad95-24ff821bf484",
      "version": 1,
      "content_hash": "sha256:54131e00af1fb617ed2983efd9a15416e4494357e2c4297251e78477f7573f91"
    },
    "export_view": "wrs-public-editorial-view-1",
    "public_view_hash": "sha256:66e44995e2dcc2d3c08b41fb6c98d488d32d46ac92e632ce527e673fecdbf568"
  },
  "target": {
    "record_id": "urn:uuid:7f4a0b2f-9007-49b7-ad95-24ff821bf484",
    "version": 1,
    "content_hash": "sha256:54131e00af1fb617ed2983efd9a15416e4494357e2c4297251e78477f7573f91"
  },
  "published_grade": {
    "profile": "wrs-published-grade-1",
    "edition": "2026-10-28.1",
    "id": "chainer-define-by-run-2019-account",
    "entry_id": "chainer-define-by-run-2019-account",
    "revision": 1,
    "title": "Chainer dynamic graphs, memory-aware autodiff and CuPy integration",
    "context": "Deep learning software",
    "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": "Documents define-by-run model construction, memory-aware reverse/double backpropagation and NumPy-like CuPy GPU integration in the Chainer framework.",
    "scope": "Documents define-by-run model construction, memory-aware reverse/double backpropagation and NumPy-like CuPy GPU integration in the Chainer framework. The 2019 account retrospectively describes the 2015 Chainer launch and the 2017 separation of CuPy. Later capabilities are not backdated. Dynamic computation and automatic differentiation have predecessors; the contribution is their define-by-run framework integration.",
    "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. The 2019 account retrospectively describes the 2015 Chainer launch and the 2017 separation of CuPy. Later capabilities are not backdated. Dynamic computation and automatic differentiation have predecessors; the contribution is their define-by-run framework integration.",
    "references": [
      {
        "title": "Chainer: A Deep Learning Framework for Accelerating the Research Cycle",
        "url": "https://arxiv.org/pdf/1908.00213v1",
        "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:7f4a0b2f-9007-49b7-ad95-24ff821bf484",
      "version": 1,
      "content_hash": "sha256:54131e00af1fb617ed2983efd9a15416e4494357e2c4297251e78477f7573f91"
    },
    "scientific_correctness_certified": false,
    "independent_expert_calibration": false,
    "method": "PROPRIETARY",
    "points": "140000"
  },
  "canonical_agent_submission": false,
  "scientific_verification_awarded": false,
  "reader_guide": null,
  "reading_status": "SOURCE_LINKED_NOTE"
}