{
  "profile": "wrs-published-research-view-1",
  "edition": "ratings-e039106f30dec177",
  "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS",
  "entry": {
    "id": "tan-le-efficientnetv2-progressive-training-2021",
    "title": "EfficientNetV2: training-aware search and progressive regularization",
    "summary": "Jointly optimized training speed and parameter efficiency through architecture search/scaling and adjusted regularization as training resolution increased.",
    "boundary": "Training-aware architecture search selected network designs within a human-defined space; researchers also supplied scaling and progressive-training choices. Inherits EfficientNet, fused mobile convolutions and progressive resizing. Delta is the training-aware integrated design and adaptive regularization, not a model-size version. Relative training gains depend on resolution, hardware, data and baseline.",
    "area": "Efficient vision training",
    "topic": "Artificial intelligence",
    "year": 2021,
    "lane": "AI",
    "result": "ALGORITHM",
    "execution_attribution": {
      "kind": "AUTOMATED_ARCHITECTURE_SEARCH",
      "statement": "Training-aware architecture search selected network designs within a human-defined space; researchers also supplied scaling and progressive-training choices.",
      "locator": "Pinned v1 abstract: training-aware neural architecture search, scaling and progressive learning.",
      "sources": [
        {
          "title": "EfficientNetV2: Smaller Models and Faster Training",
          "url": "https://arxiv.org/abs/2104.00298v1"
        }
      ],
      "attribution_basis": "SOURCE_REPORTED",
      "provider_attestation": "NONE",
      "scientific_correctness_certified": false
    },
    "package_id": "tan-le-efficientnetv2-progressive-training-2021",
    "target": {
      "record_id": "urn:uuid:8a9ede74-d794-49f2-ae56-a6fa444f4b50",
      "version": 1,
      "content_hash": "sha256:2b5fb394322a844058aa6d6f80373f41e1e9d83f51d36fa9a4f7adc32ba8c741"
    },
    "humans": [
      "Mingxing Tan",
      "Quoc V. Le"
    ],
    "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": [
      {
        "provider": "Not recorded: research execution is source-reported, not provider-attested",
        "product": "EfficientNetV2 training-aware architecture search",
        "model": "Not applicable: no language-model version identified for this search",
        "role": "Training-aware architecture search selected network designs within a human-defined space; researchers also supplied scaling and progressive-training choices.",
        "assurance": "SOURCE_REPORTED",
        "source": {
          "title": "EfficientNetV2: Smaller Models and Faster Training",
          "url": "https://arxiv.org/abs/2104.00298v1"
        }
      }
    ],
    "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-2104-00298v1",
        "title": "EfficientNetV2: Smaller Models and Faster Training",
        "url": "https://arxiv.org/abs/2104.00298v1"
      }
    ],
    "points": "120000",
    "published_grade": {
      "id": "tan-le-efficientnetv2-progressive-training-2021",
      "entry_id": "tan-le-efficientnetv2-progressive-training-2021",
      "band": "A",
      "modifier": null,
      "target": {
        "record_id": "urn:uuid:8a9ede74-d794-49f2-ae56-a6fa444f4b50",
        "version": 1,
        "content_hash": "sha256:2b5fb394322a844058aa6d6f80373f41e1e9d83f51d36fa9a4f7adc32ba8c741"
      },
      "what_changed": "Jointly optimized training speed and parameter efficiency through architecture search/scaling and adjusted regularization as training resolution increased.",
      "scope": "Jointly optimized training speed and parameter efficiency through architecture search/scaling and adjusted regularization as training resolution increased. Training-aware architecture search selected network designs within a human-defined space; researchers also supplied scaling and progressive-training choices. Inherits EfficientNet, fused mobile convolutions and progressive resizing. Delta is the training-aware integrated design and adaptive regularization, not a model-size version. Relative training gains depend on resolution, hardware, data and baseline.",
      "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:8a9ede74-d794-49f2-ae56-a6fa444f4b50",
        "version": 1,
        "content_hash": "sha256:2b5fb394322a844058aa6d6f80373f41e1e9d83f51d36fa9a4f7adc32ba8c741"
      },
      "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": "Mingxing Tan",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "EfficientNetV2: Smaller Models and Faster Training",
            "url": "https://arxiv.org/abs/2104.00298v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Quoc V. Le",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "EfficientNetV2: Smaller Models and Faster Training",
            "url": "https://arxiv.org/abs/2104.00298v1"
          }
        }
      ],
      "evidence": [
        {
          "description": "Bibliographic/stated-result source; not a WRS proof verification",
          "id": "source-0",
          "kind": "source_reference",
          "source": {
            "citation": "EfficientNetV2: Smaller Models and Faster Training",
            "url": "https://arxiv.org/abs/2104.00298v1"
          }
        }
      ],
      "origin": "AGENT_NATIVE",
      "record_profile": "wrs-local-research-notes-0.1",
      "record_type": "research_note",
      "rrp_version": "0.1",
      "scope": {
        "description": "Training-aware architecture search selected network designs within a human-defined space; researchers also supplied scaling and progressive-training choices. Inherits EfficientNet, fused mobile convolutions and progressive resizing. Delta is the training-aware integrated design and adaptive regularization, not a model-size version. Relative training gains depend on resolution, hardware, data and baseline.",
        "domain": "Efficient vision training"
      },
      "sources": [
        {
          "citation": "EfficientNetV2: Smaller Models and Faster Training",
          "url": "https://arxiv.org/abs/2104.00298v1"
        }
      ],
      "statement": "# EfficientNetV2: training-aware search and progressive regularization\n\nJointly optimized training speed and parameter efficiency through architecture search/scaling and adjusted regularization as training resolution increased.\n\n## Scientific scope\n\nTraining-aware architecture search selected network designs within a human-defined space; researchers also supplied scaling and progressive-training choices. Inherits EfficientNet, fused mobile convolutions and progressive resizing. Delta is the training-aware integrated design and adaptive regularization, not a model-size version. Relative training gains depend on resolution, hardware, data and baseline.\n\n## People and systems\n\nMingxing Tan; Quoc V. Le\n\nEfficientNetV2 training-aware architecture search / Not recorded: research execution is source-reported, not provider-attested / Not applicable: no language-model version identified for this search: Training-aware architecture search selected network designs within a human-defined space; researchers also supplied scaling and progressive-training choices. (source reported).\n\n## Original sources\n\n- [EfficientNetV2: Smaller Models and Faster Training](https://arxiv.org/abs/2104.00298v1)\n",
      "summary": "Jointly optimized training speed and parameter efficiency through architecture search/scaling and adjusted regularization as training resolution increased.",
      "title": "EfficientNetV2: training-aware search and progressive regularization"
    },
    "record_id": "urn:uuid:8a9ede74-d794-49f2-ae56-a6fa444f4b50",
    "version": 1,
    "source_ref": {
      "record_id": "urn:uuid:8a9ede74-d794-49f2-ae56-a6fa444f4b50",
      "version": 1,
      "content_hash": "sha256:2b5fb394322a844058aa6d6f80373f41e1e9d83f51d36fa9a4f7adc32ba8c741"
    },
    "export_view": "wrs-public-editorial-view-1",
    "public_view_hash": "sha256:e410584ad99e51e8d294923893f3bcb14f6b311bbf5f1e5da82a835d9c393dc6"
  },
  "target": {
    "record_id": "urn:uuid:8a9ede74-d794-49f2-ae56-a6fa444f4b50",
    "version": 1,
    "content_hash": "sha256:2b5fb394322a844058aa6d6f80373f41e1e9d83f51d36fa9a4f7adc32ba8c741"
  },
  "published_grade": {
    "profile": "wrs-published-grade-1",
    "edition": "2026-10-28.1",
    "id": "tan-le-efficientnetv2-progressive-training-2021",
    "entry_id": "tan-le-efficientnetv2-progressive-training-2021",
    "revision": 1,
    "title": "EfficientNetV2: training-aware search and progressive regularization",
    "context": "Efficient vision training",
    "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": "Jointly optimized training speed and parameter efficiency through architecture search/scaling and adjusted regularization as training resolution increased.",
    "scope": "Jointly optimized training speed and parameter efficiency through architecture search/scaling and adjusted regularization as training resolution increased. Training-aware architecture search selected network designs within a human-defined space; researchers also supplied scaling and progressive-training choices. Inherits EfficientNet, fused mobile convolutions and progressive resizing. Delta is the training-aware integrated design and adaptive regularization, not a model-size version. Relative training gains depend on resolution, hardware, data and baseline.",
    "evidence_summary": "Training-aware architecture search selected network designs within a human-defined space; researchers also supplied scaling and progressive-training choices. WRS has not independently rerun these experiments. Training-aware architecture search selected network designs within a human-defined space; researchers also supplied scaling and progressive-training choices. Inherits EfficientNet, fused mobile convolutions and progressive resizing. Delta is the training-aware integrated design and adaptive regularization, not a model-size version. Relative training gains depend on resolution, hardware, data and baseline.",
    "references": [
      {
        "title": "EfficientNetV2: Smaller Models and Faster Training",
        "url": "https://arxiv.org/abs/2104.00298v1",
        "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:8a9ede74-d794-49f2-ae56-a6fa444f4b50",
      "version": 1,
      "content_hash": "sha256:2b5fb394322a844058aa6d6f80373f41e1e9d83f51d36fa9a4f7adc32ba8c741"
    },
    "scientific_correctness_certified": false,
    "independent_expert_calibration": false,
    "method": "PROPRIETARY",
    "points": "120000"
  },
  "canonical_agent_submission": false,
  "scientific_verification_awarded": false,
  "reader_guide": null,
  "reading_status": "SOURCE_LINKED_NOTE"
}