{ "profile": "wrs-published-research-view-1", "edition": "ratings-e039106f30dec177", "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS", "entry": { "id": "perolat-et-al-deepnash-stratego-2022", "title": "DeepNash: regularized Nash dynamics for Stratego", "summary": "Combined regularized payoff-learning dynamics with model-free deep self-play, fine-tuning and test-time policy filtering to reach expert-level Stratego Classic performance without game-tree search or human demonstrations.", "boundary": "Idealized regularized-dynamics convergence and a practical deep agent's finite online ranking are different evidence. Function approximation, fine-tuning and post-processing do not constitute a WRS-certified full-game equilibrium or strong solution. Ranking concerns the reported2022Gravon sample, not an enduring all-platform world rank. Source convergence arguments inherit earlier regularization work.", "area": "Imperfect-information multiagent learning", "topic": "Artificial intelligence", "year": 2022, "lane": "HUMAN", "result": "ALGORITHM", "package_id": "perolat-et-al-deepnash-stratego-2022", "target": { "record_id": "urn:uuid:82bd155f-39ee-41f2-a8f8-b58dc5df96c9", "version": 1, "content_hash": "sha256:c4af0edce62cef6b378e5f20cc32e7822f97fa0132f632420f61726b3c2534b0" }, "humans": [ "Julien Perolat", "Bart de Vylder", "Daniel Hennes", "Eugene Tarassov", "Florian Strub", "Vincent de Boer", "Paul Muller", "Jerome T. Connor", "Neil Burch", "Thomas Anthony", "Stephen McAleer", "Romuald Elie", "Sarah H. Cen", "Zhe Wang", "Audrunas Gruslys", "Aleksandra Malysheva", "Mina Khan", "Sherjil Ozair", "Finbarr Timbers", "Toby Pohlen", "Tom Eccles", "Mark Rowland", "Marc Lanctot", "Jean-Baptiste Lespiau", "Bilal Piot", "Shayegan Omidshafiei", "Edward Lockhart", "Laurent Sifre", "Nathalie Beauguerlange", "Remi Munos", "David Silver", "Satinder Singh", "Demis Hassabis", "Karl Tuyls" ], "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-2206-15378v1", "title": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } ], "points": "155000", "published_grade": { "id": "perolat-et-al-deepnash-stratego-2022", "entry_id": "perolat-et-al-deepnash-stratego-2022", "band": "A", "modifier": null, "target": { "record_id": "urn:uuid:82bd155f-39ee-41f2-a8f8-b58dc5df96c9", "version": 1, "content_hash": "sha256:c4af0edce62cef6b378e5f20cc32e7822f97fa0132f632420f61726b3c2534b0" }, "what_changed": "Combined regularized payoff-learning dynamics with model-free deep self-play, fine-tuning and test-time policy filtering to reach expert-level Stratego Classic performance without game-tree search or human demonstrations.", "scope": "Combined regularized payoff-learning dynamics with model-free deep self-play, fine-tuning and test-time policy filtering to reach expert-level Stratego Classic performance without game-tree search or human demonstrations. Idealized regularized-dynamics convergence and a practical deep agent's finite online ranking are different evidence. Function approximation, fine-tuning and post-processing do not constitute a WRS-certified full-game equilibrium or strong solution. Ranking concerns the reported2022Gravon sample, not an enduring all-platform world rank. Source convergence arguments inherit earlier regularization work.", "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:82bd155f-39ee-41f2-a8f8-b58dc5df96c9", "version": 1, "content_hash": "sha256:c4af0edce62cef6b378e5f20cc32e7822f97fa0132f632420f61726b3c2534b0" }, "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": "Julien Perolat", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Bart de Vylder", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Daniel Hennes", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Eugene Tarassov", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Florian Strub", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Vincent de Boer", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Paul Muller", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Jerome T. Connor", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Neil Burch", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Thomas Anthony", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Stephen McAleer", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Romuald Elie", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Sarah H. Cen", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Zhe Wang", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Audrunas Gruslys", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Aleksandra Malysheva", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Mina Khan", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Sherjil Ozair", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Finbarr Timbers", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Toby Pohlen", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Tom Eccles", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Mark Rowland", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Marc Lanctot", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Jean-Baptiste Lespiau", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Bilal Piot", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Shayegan Omidshafiei", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Edward Lockhart", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Laurent Sifre", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Nathalie Beauguerlange", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "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": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "David Silver", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } } ], "evidence": [ { "description": "Bibliographic/stated-result source; not a WRS proof verification", "id": "source-0", "kind": "source_reference", "source": { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } } ], "origin": "HUMAN_LEGACY", "record_profile": "wrs-local-research-notes-0.1", "record_type": "research_note", "rrp_version": "0.1", "scope": { "description": "Idealized regularized-dynamics convergence and a practical deep agent's finite online ranking are different evidence. Function approximation, fine-tuning and post-processing do not constitute a WRS-certified full-game equilibrium or strong solution. Ranking concerns the reported2022Gravon sample, not an enduring all-platform world rank. Source convergence arguments inherit earlier regularization work.", "domain": "Imperfect-information multiagent learning" }, "sources": [ { "citation": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1" } ], "statement": "# DeepNash: regularized Nash dynamics for Stratego\n\nCombined regularized payoff-learning dynamics with model-free deep self-play, fine-tuning and test-time policy filtering to reach expert-level Stratego Classic performance without game-tree search or human demonstrations.\n\n## Scientific scope\n\nIdealized regularized-dynamics convergence and a practical deep agent's finite online ranking are different evidence. Function approximation, fine-tuning and post-processing do not constitute a WRS-certified full-game equilibrium or strong solution. Ranking concerns the reported2022Gravon sample, not an enduring all-platform world rank. Source convergence arguments inherit earlier regularization work.\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\nJulien Perolat; Bart de Vylder; Daniel Hennes; Eugene Tarassov; Florian Strub; Vincent de Boer; Paul Muller; Jerome T. Connor; Neil Burch; Thomas Anthony; Stephen McAleer; Romuald Elie; Sarah H. Cen; Zhe Wang; Audrunas Gruslys; Aleksandra Malysheva; Mina Khan; Sherjil Ozair; Finbarr Timbers; Toby Pohlen; Tom Eccles; Mark Rowland; Marc Lanctot; Jean-Baptiste Lespiau; Bilal Piot; Shayegan Omidshafiei; Edward Lockhart; Laurent Sifre; Nathalie Beauguerlange; Remi Munos; David Silver; Satinder Singh; Demis Hassabis; Karl Tuyls\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- [Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning](https://arxiv.org/pdf/2206.15378v1)\n", "summary": "Combined regularized payoff-learning dynamics with model-free deep self-play, fine-tuning and test-time policy filtering to reach expert-level Stratego Classic performance without game-tree search or human demonstrations.", "title": "DeepNash: regularized Nash dynamics for Stratego" }, "record_id": "urn:uuid:82bd155f-39ee-41f2-a8f8-b58dc5df96c9", "version": 1, "source_ref": { "record_id": "urn:uuid:82bd155f-39ee-41f2-a8f8-b58dc5df96c9", "version": 1, "content_hash": "sha256:c4af0edce62cef6b378e5f20cc32e7822f97fa0132f632420f61726b3c2534b0" }, "export_view": "wrs-public-editorial-view-1", "public_view_hash": "sha256:d46af36cdf0c5030c3199486fd5e2a2c4dfb748af73b2b9a2168831f9f8d5e35" }, "target": { "record_id": "urn:uuid:82bd155f-39ee-41f2-a8f8-b58dc5df96c9", "version": 1, "content_hash": "sha256:c4af0edce62cef6b378e5f20cc32e7822f97fa0132f632420f61726b3c2534b0" }, "published_grade": { "profile": "wrs-published-grade-1", "edition": "2026-10-28.1", "id": "perolat-et-al-deepnash-stratego-2022", "entry_id": "perolat-et-al-deepnash-stratego-2022", "revision": 1, "title": "DeepNash: regularized Nash dynamics for Stratego", "context": "Imperfect-information multiagent learning", "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": "Combined regularized payoff-learning dynamics with model-free deep self-play, fine-tuning and test-time policy filtering to reach expert-level Stratego Classic performance without game-tree search or human demonstrations.", "scope": "Combined regularized payoff-learning dynamics with model-free deep self-play, fine-tuning and test-time policy filtering to reach expert-level Stratego Classic performance without game-tree search or human demonstrations. Idealized regularized-dynamics convergence and a practical deep agent's finite online ranking are different evidence. Function approximation, fine-tuning and post-processing do not constitute a WRS-certified full-game equilibrium or strong solution. Ranking concerns the reported2022Gravon sample, not an enduring all-platform world rank. Source convergence arguments inherit earlier regularization work.", "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. Idealized regularized-dynamics convergence and a practical deep agent's finite online ranking are different evidence. Function approximation, fine-tuning and post-processing do not constitute a WRS-certified full-game equilibrium or strong solution. Ranking concerns the reported2022Gravon sample, not an enduring all-platform world rank. Source convergence arguments inherit earlier regularization work.", "references": [ { "title": "Mastering the Game of Stratego with Model-Free Multiagent Reinforcement Learning", "url": "https://arxiv.org/pdf/2206.15378v1", "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:82bd155f-39ee-41f2-a8f8-b58dc5df96c9", "version": 1, "content_hash": "sha256:c4af0edce62cef6b378e5f20cc32e7822f97fa0132f632420f61726b3c2534b0" }, "scientific_correctness_certified": false, "independent_expert_calibration": false, "method": "PROPRIETARY", "points": "155000" }, "canonical_agent_submission": false, "scientific_verification_awarded": false, "reader_guide": null, "reading_status": "SOURCE_LINKED_NOTE" }