{
  "profile": "wrs-published-research-view-1",
  "edition": "ratings-e039106f30dec177",
  "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS",
  "entry": {
    "id": "abdolmaleki-song-et-al-mpo-vmpo-family-2018",
    "title": "MPO and V-MPO: constrained target fitting for policy improvement",
    "summary": "Used an expectation-maximization style policy-improvement step with constrained nonparametric targets followed by parametric policy fitting; V-MPO adapts this approach to on-policy, value-based advantage weighting.",
    "boundary": "Bundled off-policy MPO2018 and on-policy V-MPO2019 with distinct original bylines, not one common author list. Inherits EM, relative-entropy policy search and policy iteration; KL constraints and reported stability do not ensure globally optimal neural policies or arbitrary trust-region equivalence.",
    "area": "Policy optimization methods",
    "topic": "Artificial intelligence",
    "year": 2018,
    "lane": "HUMAN",
    "result": "ALGORITHM",
    "package_id": "abdolmaleki-song-et-al-mpo-vmpo-family-2018",
    "target": {
      "record_id": "urn:uuid:d3a58a39-b9f8-4812-a567-982ff6ab0d54",
      "version": 1,
      "content_hash": "sha256:6d60b4407320ab9953997f0b6e5cdea7b0d4ae67cb00e7b9a6a2862c71812dc1"
    },
    "humans": [
      "Abbas Abdolmaleki",
      "Jost Tobias Springenberg",
      "Yuval Tassa",
      "Remi Munos",
      "Nicolas Heess",
      "Martin Riedmiller",
      "H. Francis Song",
      "Aidan Clark",
      "Hubert Soyer",
      "Jack W. Rae",
      "Seb Noury",
      "Arun Ahuja",
      "Siqi Liu",
      "Dhruva Tirumala",
      "Dan Belov",
      "Matthew M. Botvinick"
    ],
    "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-1806-06920v1",
        "title": "Maximum a Posteriori Policy Optimisation",
        "url": "https://arxiv.org/abs/1806.06920v1"
      },
      {
        "id": "cur20261015-arxiv-1909-12238v1",
        "title": "V-MPO: On-Policy Maximum a Posteriori Policy Optimization for Discrete and Continuous Control",
        "url": "https://arxiv.org/abs/1909.12238v1"
      }
    ],
    "points": "115000",
    "published_grade": {
      "id": "abdolmaleki-song-et-al-mpo-vmpo-family-2018",
      "entry_id": "abdolmaleki-song-et-al-mpo-vmpo-family-2018",
      "band": "A",
      "modifier": null,
      "target": {
        "record_id": "urn:uuid:d3a58a39-b9f8-4812-a567-982ff6ab0d54",
        "version": 1,
        "content_hash": "sha256:6d60b4407320ab9953997f0b6e5cdea7b0d4ae67cb00e7b9a6a2862c71812dc1"
      },
      "what_changed": "Used an expectation-maximization style policy-improvement step with constrained nonparametric targets followed by parametric policy fitting; V-MPO adapts this approach to on-policy, value-based advantage weighting.",
      "scope": "Used an expectation-maximization style policy-improvement step with constrained nonparametric targets followed by parametric policy fitting; V-MPO adapts this approach to on-policy, value-based advantage weighting. Bundled off-policy MPO2018 and on-policy V-MPO2019 with distinct original bylines, not one common author list. Inherits EM, relative-entropy policy search and policy iteration; KL constraints and reported stability do not ensure globally optimal neural policies or arbitrary trust-region equivalence.",
      "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:d3a58a39-b9f8-4812-a567-982ff6ab0d54",
        "version": 1,
        "content_hash": "sha256:6d60b4407320ab9953997f0b6e5cdea7b0d4ae67cb00e7b9a6a2862c71812dc1"
      },
      "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": "Abbas Abdolmaleki",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Maximum a Posteriori Policy Optimisation",
            "url": "https://arxiv.org/abs/1806.06920v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Jost Tobias Springenberg",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Maximum a Posteriori Policy Optimisation",
            "url": "https://arxiv.org/abs/1806.06920v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Yuval Tassa",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Maximum a Posteriori Policy Optimisation",
            "url": "https://arxiv.org/abs/1806.06920v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Remi Munos",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Maximum a Posteriori Policy Optimisation",
            "url": "https://arxiv.org/abs/1806.06920v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Nicolas Heess",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Maximum a Posteriori Policy Optimisation",
            "url": "https://arxiv.org/abs/1806.06920v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Martin Riedmiller",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "Maximum a Posteriori Policy Optimisation",
            "url": "https://arxiv.org/abs/1806.06920v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "H. Francis Song",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "V-MPO: On-Policy Maximum a Posteriori Policy Optimization for Discrete and Continuous Control",
            "url": "https://arxiv.org/abs/1909.12238v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Aidan Clark",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "V-MPO: On-Policy Maximum a Posteriori Policy Optimization for Discrete and Continuous Control",
            "url": "https://arxiv.org/abs/1909.12238v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Hubert Soyer",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "V-MPO: On-Policy Maximum a Posteriori Policy Optimization for Discrete and Continuous Control",
            "url": "https://arxiv.org/abs/1909.12238v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Jack W. Rae",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "V-MPO: On-Policy Maximum a Posteriori Policy Optimization for Discrete and Continuous Control",
            "url": "https://arxiv.org/abs/1909.12238v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Seb Noury",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "V-MPO: On-Policy Maximum a Posteriori Policy Optimization for Discrete and Continuous Control",
            "url": "https://arxiv.org/abs/1909.12238v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Arun Ahuja",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "V-MPO: On-Policy Maximum a Posteriori Policy Optimization for Discrete and Continuous Control",
            "url": "https://arxiv.org/abs/1909.12238v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Siqi Liu",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "V-MPO: On-Policy Maximum a Posteriori Policy Optimization for Discrete and Continuous Control",
            "url": "https://arxiv.org/abs/1909.12238v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Dhruva Tirumala",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "V-MPO: On-Policy Maximum a Posteriori Policy Optimization for Discrete and Continuous Control",
            "url": "https://arxiv.org/abs/1909.12238v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Dan Belov",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "V-MPO: On-Policy Maximum a Posteriori Policy Optimization for Discrete and Continuous Control",
            "url": "https://arxiv.org/abs/1909.12238v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Matthew M. Botvinick",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "V-MPO: On-Policy Maximum a Posteriori Policy Optimization for Discrete and Continuous Control",
            "url": "https://arxiv.org/abs/1909.12238v1"
          }
        }
      ],
      "evidence": [
        {
          "description": "Bibliographic/stated-result source; not a WRS proof verification",
          "id": "source-0",
          "kind": "source_reference",
          "source": {
            "citation": "Maximum a Posteriori Policy Optimisation",
            "url": "https://arxiv.org/abs/1806.06920v1"
          }
        },
        {
          "description": "Bibliographic/stated-result source; not a WRS proof verification",
          "id": "source-1",
          "kind": "source_reference",
          "source": {
            "citation": "V-MPO: On-Policy Maximum a Posteriori Policy Optimization for Discrete and Continuous Control",
            "url": "https://arxiv.org/abs/1909.12238v1"
          }
        }
      ],
      "origin": "HUMAN_LEGACY",
      "record_profile": "wrs-local-research-notes-0.1",
      "record_type": "research_note",
      "rrp_version": "0.1",
      "scope": {
        "description": "Bundled off-policy MPO2018 and on-policy V-MPO2019 with distinct original bylines, not one common author list. Inherits EM, relative-entropy policy search and policy iteration; KL constraints and reported stability do not ensure globally optimal neural policies or arbitrary trust-region equivalence.",
        "domain": "Policy optimization methods"
      },
      "sources": [
        {
          "citation": "Maximum a Posteriori Policy Optimisation",
          "url": "https://arxiv.org/abs/1806.06920v1"
        },
        {
          "citation": "V-MPO: On-Policy Maximum a Posteriori Policy Optimization for Discrete and Continuous Control",
          "url": "https://arxiv.org/abs/1909.12238v1"
        }
      ],
      "statement": "# MPO and V-MPO: constrained target fitting for policy improvement\n\nUsed an expectation-maximization style policy-improvement step with constrained nonparametric targets followed by parametric policy fitting; V-MPO adapts this approach to on-policy, value-based advantage weighting.\n\n## Scientific scope\n\nBundled off-policy MPO2018 and on-policy V-MPO2019 with distinct original bylines, not one common author list. Inherits EM, relative-entropy policy search and policy iteration; KL constraints and reported stability do not ensure globally optimal neural policies or arbitrary trust-region equivalence.\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\nAbbas Abdolmaleki; Jost Tobias Springenberg; Yuval Tassa; Remi Munos; Nicolas Heess; Martin Riedmiller; H. Francis Song; Aidan Clark; Hubert Soyer; Jack W. Rae; Seb Noury; Arun Ahuja; Siqi Liu; Dhruva Tirumala; Dan Belov; Matthew M. Botvinick\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- [Maximum a Posteriori Policy Optimisation](https://arxiv.org/abs/1806.06920v1)\n- [V-MPO: On-Policy Maximum a Posteriori Policy Optimization for Discrete and Continuous Control](https://arxiv.org/abs/1909.12238v1)\n",
      "summary": "Used an expectation-maximization style policy-improvement step with constrained nonparametric targets followed by parametric policy fitting; V-MPO adapts this approach to on-policy, value-based advantage weighting.",
      "title": "MPO and V-MPO: constrained target fitting for policy improvement"
    },
    "record_id": "urn:uuid:d3a58a39-b9f8-4812-a567-982ff6ab0d54",
    "version": 1,
    "source_ref": {
      "record_id": "urn:uuid:d3a58a39-b9f8-4812-a567-982ff6ab0d54",
      "version": 1,
      "content_hash": "sha256:6d60b4407320ab9953997f0b6e5cdea7b0d4ae67cb00e7b9a6a2862c71812dc1"
    },
    "export_view": "wrs-public-editorial-view-1",
    "public_view_hash": "sha256:bcdbc6c747ceb89533893e2d9d4eee6bfd4b89f160275479e1eb76142e3232dd"
  },
  "target": {
    "record_id": "urn:uuid:d3a58a39-b9f8-4812-a567-982ff6ab0d54",
    "version": 1,
    "content_hash": "sha256:6d60b4407320ab9953997f0b6e5cdea7b0d4ae67cb00e7b9a6a2862c71812dc1"
  },
  "published_grade": {
    "profile": "wrs-published-grade-1",
    "edition": "2026-10-28.1",
    "id": "abdolmaleki-song-et-al-mpo-vmpo-family-2018",
    "entry_id": "abdolmaleki-song-et-al-mpo-vmpo-family-2018",
    "revision": 1,
    "title": "MPO and V-MPO: constrained target fitting for policy improvement",
    "context": "Policy optimization methods",
    "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": "Used an expectation-maximization style policy-improvement step with constrained nonparametric targets followed by parametric policy fitting; V-MPO adapts this approach to on-policy, value-based advantage weighting.",
    "scope": "Used an expectation-maximization style policy-improvement step with constrained nonparametric targets followed by parametric policy fitting; V-MPO adapts this approach to on-policy, value-based advantage weighting. Bundled off-policy MPO2018 and on-policy V-MPO2019 with distinct original bylines, not one common author list. Inherits EM, relative-entropy policy search and policy iteration; KL constraints and reported stability do not ensure globally optimal neural policies or arbitrary trust-region equivalence.",
    "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. Bundled off-policy MPO2018 and on-policy V-MPO2019 with distinct original bylines, not one common author list. Inherits EM, relative-entropy policy search and policy iteration; KL constraints and reported stability do not ensure globally optimal neural policies or arbitrary trust-region equivalence.",
    "references": [
      {
        "title": "Maximum a Posteriori Policy Optimisation",
        "url": "https://arxiv.org/abs/1806.06920v1",
        "locator": "Original contribution and stated assumptions, evaluation scope and limitations; source-linked assessment, not independent execution or proof certification."
      },
      {
        "title": "V-MPO: On-Policy Maximum a Posteriori Policy Optimization for Discrete and Continuous Control",
        "url": "https://arxiv.org/abs/1909.12238v1",
        "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:d3a58a39-b9f8-4812-a567-982ff6ab0d54",
      "version": 1,
      "content_hash": "sha256:6d60b4407320ab9953997f0b6e5cdea7b0d4ae67cb00e7b9a6a2862c71812dc1"
    },
    "scientific_correctness_certified": false,
    "independent_expert_calibration": false,
    "method": "PROPRIETARY",
    "points": "115000"
  },
  "canonical_agent_submission": false,
  "scientific_verification_awarded": false,
  "reader_guide": null,
  "reading_status": "SOURCE_LINKED_NOTE"
}