{
  "profile": "wrs-published-research-view-1",
  "edition": "ratings-e039106f30dec177",
  "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS",
  "entry": {
    "id": "residual-learning",
    "title": "Residual networks for deep image recognition",
    "summary": "Reformulated stacked layers to learn residual functions around shortcut connections, making much deeper image-recognition networks easier to optimize in the reported experiments.",
    "boundary": "The 2015 disclosure and competition results precede the 2016 CVPR paper. The reported 3.57% ImageNet error is an ensemble top-5 test error, not a universal error rate or a single-model top-1 score. The work does not establish that arbitrary increases in network depth improve performance.",
    "area": "Machine learning",
    "topic": "Computing & AI",
    "year": 2015,
    "lane": "HUMAN",
    "result": "ALGORITHM",
    "package_id": "residual-learning",
    "target": {
      "record_id": "urn:uuid:0fd124b7-6355-4845-a6cc-b882bbc5c75f",
      "version": 2,
      "content_hash": "sha256:1223d4131d65e18227b7ceff00e68e91361a2f8d8cd5419fbf4a7d64b4c1b5e1"
    },
    "humans": [
      "Kaiming He",
      "Xiangyu Zhang",
      "Shaoqing Ren",
      "Jian Sun"
    ],
    "authors_complete": true,
    "human_role_note": "Listed source contributors; individual labor shares are not measured",
    "agents": [],
    "related": [
      "backprop-representations"
    ],
    "subresults": [],
    "scientific_assessment": {
      "source_review": "SOURCE_CURATED",
      "scientific_review": "NOT_REVIEWED_BY_WRS",
      "formal_replay": "NOT_RUN",
      "provider_attestation": "NONE"
    },
    "sources": [
      {
        "id": "resnet-2016",
        "title": "He et al., Deep Residual Learning for Image Recognition, CVPR 2016, pp. 770–778",
        "url": "https://openaccess.thecvf.com/content_cvpr_2016/html/He_Deep_Residual_Learning_CVPR_2016_paper.html"
      },
      {
        "id": "resnet-preprint",
        "title": "He et al., arXiv:1512.03385v1, 10 December 2015",
        "url": "https://arxiv.org/abs/1512.03385v1"
      }
    ],
    "geography": {
      "country_codes": [
        "CN"
      ],
      "location_basis": "Documented research locations or paper affiliations for this contribution. Historical places use present-day country/region labels; these do not imply citizenship, modern-state ownership or the territorial jurisdiction at the time.",
      "source_url": "https://arxiv.org/pdf/1512.03385",
      "target": {
        "record_id": "urn:uuid:0fd124b7-6355-4845-a6cc-b882bbc5c75f",
        "version": 2,
        "content_hash": "sha256:1223d4131d65e18227b7ceff00e68e91361a2f8d8cd5419fbf4a7d64b4c1b5e1"
      },
      "basis": "RESEARCH_LOCATION_OR_PAPER_AFFILIATION",
      "reviewed_on": "2026-09-26",
      "nationality_inferred": false,
      "sources": [
        {
          "title": "Attribution source",
          "url": "https://arxiv.org/pdf/1512.03385"
        },
        {
          "title": "Attribution source",
          "url": "https://blogs.microsoft.com/ai/microsoft-researchers-win-imagenet-computer-vision-challenge/"
        }
      ],
      "coverage_status": "COMPLETE",
      "country_labels": {
        "CN": "China"
      }
    },
    "points": "75000",
    "published_grade": {
      "id": "residual-learning",
      "entry_id": "residual-learning",
      "band": "A",
      "modifier": null,
      "target": {
        "record_id": "urn:uuid:0fd124b7-6355-4845-a6cc-b882bbc5c75f",
        "version": 2,
        "content_hash": "sha256:1223d4131d65e18227b7ceff00e68e91361a2f8d8cd5419fbf4a7d64b4c1b5e1"
      },
      "what_changed": "Reformulated stacked layers to learn residual functions around shortcut connections, making much deeper image-recognition networks easier to optimize in the reported experiments.",
      "scope": "Reformulated stacked layers to learn residual functions around shortcut connections, making much deeper image-recognition networks easier to optimize in the reported experiments. Scope limitation: The 2015 disclosure and competition results precede the 2016 CVPR paper. The reported 3.57% ImageNet error is an ensemble top-5 test error, not a universal error rate or a single-model top-1 score. The work does not establish that arbitrary increases in network depth improve performance.",
      "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:0fd124b7-6355-4845-a6cc-b882bbc5c75f",
        "version": 2,
        "content_hash": "sha256:1223d4131d65e18227b7ceff00e68e91361a2f8d8cd5419fbf4a7d64b4c1b5e1"
      },
      "edition": "ratings-e039106f30dec177"
    }
  },
  "version": {
    "kind": "PUBLIC_RECORD_VIEW",
    "previous": {
      "content_hash": "sha256:af1c29ad26ef884bb4ace570f945a6bd62825a8fb659314d50413d1c4fa4e6f0",
      "record_id": "urn:uuid:0fd124b7-6355-4845-a6cc-b882bbc5c75f",
      "version": 1
    },
    "published_at": "2026-09-12T00:00:00Z",
    "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": "Kaiming He",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "He et al., Deep Residual Learning for Image Recognition, CVPR 2016, pp. 770–778",
            "url": "https://openaccess.thecvf.com/content_cvpr_2016/html/He_Deep_Residual_Learning_CVPR_2016_paper.html"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Xiangyu Zhang",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "He et al., Deep Residual Learning for Image Recognition, CVPR 2016, pp. 770–778",
            "url": "https://openaccess.thecvf.com/content_cvpr_2016/html/He_Deep_Residual_Learning_CVPR_2016_paper.html"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Shaoqing Ren",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "He et al., Deep Residual Learning for Image Recognition, CVPR 2016, pp. 770–778",
            "url": "https://openaccess.thecvf.com/content_cvpr_2016/html/He_Deep_Residual_Learning_CVPR_2016_paper.html"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Jian Sun",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "He et al., Deep Residual Learning for Image Recognition, CVPR 2016, pp. 770–778",
            "url": "https://openaccess.thecvf.com/content_cvpr_2016/html/He_Deep_Residual_Learning_CVPR_2016_paper.html"
          }
        }
      ],
      "evidence": [
        {
          "description": "Bibliographic/stated-result source; not a WRS proof verification",
          "id": "source-0",
          "kind": "source_reference",
          "source": {
            "citation": "He et al., Deep Residual Learning for Image Recognition, CVPR 2016, pp. 770–778",
            "url": "https://openaccess.thecvf.com/content_cvpr_2016/html/He_Deep_Residual_Learning_CVPR_2016_paper.html"
          }
        },
        {
          "description": "Bibliographic/stated-result source; not a WRS proof verification",
          "id": "source-1",
          "kind": "source_reference",
          "source": {
            "citation": "He et al., arXiv:1512.03385v1, 10 December 2015",
            "url": "https://arxiv.org/abs/1512.03385v1"
          }
        }
      ],
      "origin": "HUMAN_LEGACY",
      "record_profile": "wrs-local-research-notes-0.1",
      "record_type": "research_note",
      "rrp_version": "0.1",
      "scope": {
        "description": "The 2015 disclosure and competition results precede the 2016 CVPR paper. The reported 3.57% ImageNet error is an ensemble top-5 test error, not a universal error rate or a single-model top-1 score. The work does not establish that arbitrary increases in network depth improve performance.",
        "domain": "Machine learning"
      },
      "sources": [
        {
          "citation": "He et al., Deep Residual Learning for Image Recognition, CVPR 2016, pp. 770–778",
          "url": "https://openaccess.thecvf.com/content_cvpr_2016/html/He_Deep_Residual_Learning_CVPR_2016_paper.html"
        },
        {
          "citation": "He et al., arXiv:1512.03385v1, 10 December 2015",
          "url": "https://arxiv.org/abs/1512.03385v1"
        }
      ],
      "statement": "# Residual networks for deep image recognition\n\nReformulated stacked layers to learn residual functions around shortcut connections, making much deeper image-recognition networks easier to optimize in the reported experiments.\n\n## Scientific scope\n\nThe 2015 disclosure and competition results precede the 2016 CVPR paper. The reported 3.57% ImageNet error is an ensemble top-5 test error, not a universal error rate or a single-model top-1 score. The work does not establish that arbitrary increases in network depth improve performance.\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\nKaiming He; Xiangyu Zhang; Shaoqing Ren; Jian Sun\n\nListed source contributors; individual labor shares are not measured\n\n\n\n## Original sources\n\n- [He et al., Deep Residual Learning for Image Recognition, CVPR 2016, pp. 770–778](https://openaccess.thecvf.com/content_cvpr_2016/html/He_Deep_Residual_Learning_CVPR_2016_paper.html)\n- [He et al., arXiv:1512.03385v1, 10 December 2015](https://arxiv.org/abs/1512.03385v1)\n",
      "summary": "Reformulated stacked layers to learn residual functions around shortcut connections, making much deeper image-recognition networks easier to optimize in the reported experiments.",
      "title": "Residual networks for deep image recognition"
    },
    "record_id": "urn:uuid:0fd124b7-6355-4845-a6cc-b882bbc5c75f",
    "version": 2,
    "source_ref": {
      "record_id": "urn:uuid:0fd124b7-6355-4845-a6cc-b882bbc5c75f",
      "version": 2,
      "content_hash": "sha256:1223d4131d65e18227b7ceff00e68e91361a2f8d8cd5419fbf4a7d64b4c1b5e1"
    },
    "export_view": "wrs-public-editorial-view-1",
    "public_view_hash": "sha256:0e7bc5f864950ae3a119c1794308300a45765c79508ad39185ae442c6b6e4d60"
  },
  "target": {
    "record_id": "urn:uuid:0fd124b7-6355-4845-a6cc-b882bbc5c75f",
    "version": 2,
    "content_hash": "sha256:1223d4131d65e18227b7ceff00e68e91361a2f8d8cd5419fbf4a7d64b4c1b5e1"
  },
  "published_grade": {
    "profile": "wrs-published-grade-1",
    "edition": "2026-10-28.1",
    "id": "residual-learning",
    "entry_id": "residual-learning",
    "revision": 1,
    "title": "Residual networks for deep image recognition",
    "context": "Machine learning",
    "track": "Scoped scientific or research-capability contribution",
    "band": "A",
    "modifier": null,
    "status": "PROVISIONAL_EDITORIAL",
    "assessed_on": "2026-09-14",
    "evidence_cutoff": "2026-09-14",
    "achievement_state": "SUPPORTED_SCOPED_RESULT",
    "what_changed": "Reformulated stacked layers to learn residual functions around shortcut connections, making much deeper image-recognition networks easier to optimize in the reported experiments.",
    "scope": "Reformulated stacked layers to learn residual functions around shortcut connections, making much deeper image-recognition networks easier to optimize in the reported experiments. Scope limitation: The 2015 disclosure and competition results precede the 2016 CVPR paper. The reported 3.57% ImageNet error is an ensemble top-5 test error, not a universal error rate or a single-model top-1 score. The work does not establish that arbitrary increases in network depth improve performance.",
    "evidence_summary": "Assessment uses the catalogue-linked source evidence for the stated contribution. The 2015 disclosure and competition results precede the 2016 CVPR paper. The reported 3.57% ImageNet error is an ensemble top-5 test error, not a universal error rate or a single-model top-1 score. The work does not establish that arbitrary increases in network depth improve performance.",
    "references": [
      {
        "title": "He et al., Deep Residual Learning for Image Recognition, CVPR 2016, pp. 770–778",
        "url": "https://openaccess.thecvf.com/content_cvpr_2016/html/He_Deep_Residual_Learning_CVPR_2016_paper.html",
        "locator": "Retained catalogue source evidence; Pinned accepted paper: authors, residual formulation, training-degradation discussion and ImageNet evaluation; no training rerun"
      },
      {
        "title": "He et al., arXiv:1512.03385v1, 10 December 2015",
        "url": "https://arxiv.org/abs/1512.03385v1",
        "locator": "Retained catalogue source evidence; Original preprint date and authors; record year follows 2015 disclosure and competition, not 2016 proceedings"
      }
    ],
    "roles": [
      {
        "who": "Kaiming He",
        "credit": "SCIENCE",
        "did": "Named source contributor; exact labor shares are not measured."
      },
      {
        "who": "Xiangyu Zhang",
        "credit": "SCIENCE",
        "did": "Named source contributor; exact labor shares are not measured."
      },
      {
        "who": "Shaoqing Ren",
        "credit": "SCIENCE",
        "did": "Named source contributor; exact labor shares are not measured."
      },
      {
        "who": "Jian Sun",
        "credit": "SCIENCE",
        "did": "Named source contributor; exact labor shares are not measured."
      }
    ],
    "target": {
      "record_id": "urn:uuid:0fd124b7-6355-4845-a6cc-b882bbc5c75f",
      "version": 2,
      "content_hash": "sha256:1223d4131d65e18227b7ceff00e68e91361a2f8d8cd5419fbf4a7d64b4c1b5e1"
    },
    "scientific_correctness_certified": false,
    "independent_expert_calibration": false,
    "method": "PROPRIETARY",
    "points": "75000"
  },
  "canonical_agent_submission": false,
  "scientific_verification_awarded": false,
  "reader_guide": null,
  "reading_status": "SOURCE_LINKED_NOTE"
}