{
  "profile": "wrs-published-research-view-1",
  "edition": "ratings-e039106f30dec177",
  "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS",
  "entry": {
    "id": "hessel-et-al-rainbow-rl-integration-2017",
    "title": "Hessel et al.: Rainbow integration and ablation of DQN improvements",
    "summary": "Integrated six complementary DQN extensions and used ablations to characterize their joint effects on Atari learning efficiency and final performance.",
    "boundary": "Double Q, prioritized replay, dueling networks, multistep returns, distributional targets and noisy exploration retain original credit. This is integration/interaction evidence, not six new inventions. Preprint2017 and AAAI publication2018 are distinct dates.",
    "area": "Deep reinforcement-learning integration",
    "topic": "Artificial intelligence",
    "year": 2017,
    "lane": "HUMAN",
    "result": "EMPIRICAL",
    "package_id": "hessel-et-al-rainbow-rl-integration-2017",
    "target": {
      "record_id": "urn:uuid:6bf73ada-dd35-4d38-a8eb-b679badf33da",
      "version": 1,
      "content_hash": "sha256:71f18659afa19920e07586e9735bebbb8e28b77a387b6abea21fb98d7a0a2fee"
    },
    "humans": [
      "Matteo Hessel",
      "Joseph Modayil",
      "Hado van Hasselt",
      "Tom Schaul",
      "Georg Ostrovski",
      "Will Dabney",
      "Dan Horgan",
      "Bilal Piot",
      "Mohammad Azar",
      "David Silver"
    ],
    "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-1710-02298v1",
        "title": "Rainbow: Combining Improvements in Deep Reinforcement Learning",
        "url": "https://arxiv.org/abs/1710.02298v1"
      }
    ],
    "points": "130000",
    "published_grade": {
      "id": "hessel-et-al-rainbow-rl-integration-2017",
      "entry_id": "hessel-et-al-rainbow-rl-integration-2017",
      "band": "A",
      "modifier": null,
      "target": {
        "record_id": "urn:uuid:6bf73ada-dd35-4d38-a8eb-b679badf33da",
        "version": 1,
        "content_hash": "sha256:71f18659afa19920e07586e9735bebbb8e28b77a387b6abea21fb98d7a0a2fee"
      },
      "what_changed": "Integrated six complementary DQN extensions and used ablations to characterize their joint effects on Atari learning efficiency and final performance.",
      "scope": "Integrated six complementary DQN extensions and used ablations to characterize their joint effects on Atari learning efficiency and final performance. Double Q, prioritized replay, dueling networks, multistep returns, distributional targets and noisy exploration retain original credit. This is integration/interaction evidence, not six new inventions. Preprint2017 and AAAI publication2018 are distinct dates.",
      "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:6bf73ada-dd35-4d38-a8eb-b679badf33da",
        "version": 1,
        "content_hash": "sha256:71f18659afa19920e07586e9735bebbb8e28b77a387b6abea21fb98d7a0a2fee"
      },
      "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": "Matteo Hessel",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Rainbow: Combining Improvements in Deep Reinforcement Learning",
            "url": "https://arxiv.org/abs/1710.02298v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Joseph Modayil",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Rainbow: Combining Improvements in Deep Reinforcement Learning",
            "url": "https://arxiv.org/abs/1710.02298v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Hado van Hasselt",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Rainbow: Combining Improvements in Deep Reinforcement Learning",
            "url": "https://arxiv.org/abs/1710.02298v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Tom Schaul",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Rainbow: Combining Improvements in Deep Reinforcement Learning",
            "url": "https://arxiv.org/abs/1710.02298v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Georg Ostrovski",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Rainbow: Combining Improvements in Deep Reinforcement Learning",
            "url": "https://arxiv.org/abs/1710.02298v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Will Dabney",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Rainbow: Combining Improvements in Deep Reinforcement Learning",
            "url": "https://arxiv.org/abs/1710.02298v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Dan Horgan",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Rainbow: Combining Improvements in Deep Reinforcement Learning",
            "url": "https://arxiv.org/abs/1710.02298v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Bilal Piot",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Rainbow: Combining Improvements in Deep Reinforcement Learning",
            "url": "https://arxiv.org/abs/1710.02298v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Mohammad Azar",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Rainbow: Combining Improvements in Deep Reinforcement Learning",
            "url": "https://arxiv.org/abs/1710.02298v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "David Silver",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "discoverer"
          ],
          "source": {
            "citation": "Rainbow: Combining Improvements in Deep Reinforcement Learning",
            "url": "https://arxiv.org/abs/1710.02298v1"
          }
        }
      ],
      "evidence": [
        {
          "description": "Bibliographic/stated-result source; not a WRS proof verification",
          "id": "source-0",
          "kind": "source_reference",
          "source": {
            "citation": "Rainbow: Combining Improvements in Deep Reinforcement Learning",
            "url": "https://arxiv.org/abs/1710.02298v1"
          }
        }
      ],
      "origin": "HUMAN_LEGACY",
      "record_profile": "wrs-local-research-notes-0.1",
      "record_type": "research_note",
      "rrp_version": "0.1",
      "scope": {
        "description": "Double Q, prioritized replay, dueling networks, multistep returns, distributional targets and noisy exploration retain original credit. This is integration/interaction evidence, not six new inventions. Preprint2017 and AAAI publication2018 are distinct dates.",
        "domain": "Deep reinforcement-learning integration"
      },
      "sources": [
        {
          "citation": "Rainbow: Combining Improvements in Deep Reinforcement Learning",
          "url": "https://arxiv.org/abs/1710.02298v1"
        }
      ],
      "statement": "# Hessel et al.: Rainbow integration and ablation of DQN improvements\n\nIntegrated six complementary DQN extensions and used ablations to characterize their joint effects on Atari learning efficiency and final performance.\n\n## Scientific scope\n\nDouble Q, prioritized replay, dueling networks, multistep returns, distributional targets and noisy exploration retain original credit. This is integration/interaction evidence, not six new inventions. Preprint2017 and AAAI publication2018 are distinct dates.\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\nMatteo Hessel; Joseph Modayil; Hado van Hasselt; Tom Schaul; Georg Ostrovski; Will Dabney; Dan Horgan; Bilal Piot; Mohammad Azar; David Silver\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- [Rainbow: Combining Improvements in Deep Reinforcement Learning](https://arxiv.org/abs/1710.02298v1)\n",
      "summary": "Integrated six complementary DQN extensions and used ablations to characterize their joint effects on Atari learning efficiency and final performance.",
      "title": "Hessel et al.: Rainbow integration and ablation of DQN improvements"
    },
    "record_id": "urn:uuid:6bf73ada-dd35-4d38-a8eb-b679badf33da",
    "version": 1,
    "source_ref": {
      "record_id": "urn:uuid:6bf73ada-dd35-4d38-a8eb-b679badf33da",
      "version": 1,
      "content_hash": "sha256:71f18659afa19920e07586e9735bebbb8e28b77a387b6abea21fb98d7a0a2fee"
    },
    "export_view": "wrs-public-editorial-view-1",
    "public_view_hash": "sha256:e3452287342a0e0582d646334026c7b8c3b34fe9a75320072b98fe244a065789"
  },
  "target": {
    "record_id": "urn:uuid:6bf73ada-dd35-4d38-a8eb-b679badf33da",
    "version": 1,
    "content_hash": "sha256:71f18659afa19920e07586e9735bebbb8e28b77a387b6abea21fb98d7a0a2fee"
  },
  "published_grade": {
    "profile": "wrs-published-grade-1",
    "edition": "2026-10-28.1",
    "id": "hessel-et-al-rainbow-rl-integration-2017",
    "entry_id": "hessel-et-al-rainbow-rl-integration-2017",
    "revision": 1,
    "title": "Hessel et al.: Rainbow integration and ablation of DQN improvements",
    "context": "Deep reinforcement-learning integration",
    "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": "Integrated six complementary DQN extensions and used ablations to characterize their joint effects on Atari learning efficiency and final performance.",
    "scope": "Integrated six complementary DQN extensions and used ablations to characterize their joint effects on Atari learning efficiency and final performance. Double Q, prioritized replay, dueling networks, multistep returns, distributional targets and noisy exploration retain original credit. This is integration/interaction evidence, not six new inventions. Preprint2017 and AAAI publication2018 are distinct dates.",
    "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. Double Q, prioritized replay, dueling networks, multistep returns, distributional targets and noisy exploration retain original credit. This is integration/interaction evidence, not six new inventions. Preprint2017 and AAAI publication2018 are distinct dates.",
    "references": [
      {
        "title": "Rainbow: Combining Improvements in Deep Reinforcement Learning",
        "url": "https://arxiv.org/abs/1710.02298v1",
        "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:6bf73ada-dd35-4d38-a8eb-b679badf33da",
      "version": 1,
      "content_hash": "sha256:71f18659afa19920e07586e9735bebbb8e28b77a387b6abea21fb98d7a0a2fee"
    },
    "scientific_correctness_certified": false,
    "independent_expert_calibration": false,
    "method": "PROPRIETARY",
    "points": "130000"
  },
  "canonical_agent_submission": false,
  "scientific_verification_awarded": false,
  "reader_guide": null,
  "reading_status": "SOURCE_LINKED_NOTE"
}