{ "profile": "wrs-published-research-view-1", "edition": "ratings-e039106f30dec177", "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS", "entry": { "id": "horgan-et-al-apex-distributed-replay-2018", "title": "Ape-X: distributed actors linked through prioritized replay", "summary": "Decoupled many diverse-exploration actors from a learner through shared prioritized replay, emphasizing high-throughput experience selection rather than actor-gradient synchronization.", "boundary": "Combines inherited prioritized replay and off-policy value learning; not their invention or the first distributed RL system. Scale and Atari/control gains are source experiments with hardware and data costs, not universal sample-efficiency or convergence guarantees.", "area": "Distributed reinforcement-learning systems", "topic": "Artificial intelligence", "year": 2018, "lane": "HUMAN", "result": "ALGORITHM", "package_id": "horgan-et-al-apex-distributed-replay-2018", "target": { "record_id": "urn:uuid:c1fb97c7-4255-4a6b-a60e-ff0e4415a7e9", "version": 1, "content_hash": "sha256:aa9fedc7400437ba85c0bb4806cc92e682b97dfdc02f340be12b910e115aafeb" }, "humans": [ "Dan Horgan", "John Quan", "David Budden", "Gabriel Barth-Maron", "Matteo Hessel", "Hado van Hasselt", "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-1803-00933v1", "title": "Distributed Prioritized Experience Replay", "url": "https://arxiv.org/abs/1803.00933v1" } ], "points": "130000", "published_grade": { "id": "horgan-et-al-apex-distributed-replay-2018", "entry_id": "horgan-et-al-apex-distributed-replay-2018", "band": "A", "modifier": null, "target": { "record_id": "urn:uuid:c1fb97c7-4255-4a6b-a60e-ff0e4415a7e9", "version": 1, "content_hash": "sha256:aa9fedc7400437ba85c0bb4806cc92e682b97dfdc02f340be12b910e115aafeb" }, "what_changed": "Decoupled many diverse-exploration actors from a learner through shared prioritized replay, emphasizing high-throughput experience selection rather than actor-gradient synchronization.", "scope": "Decoupled many diverse-exploration actors from a learner through shared prioritized replay, emphasizing high-throughput experience selection rather than actor-gradient synchronization. Combines inherited prioritized replay and off-policy value learning; not their invention or the first distributed RL system. Scale and Atari/control gains are source experiments with hardware and data costs, not universal sample-efficiency or convergence guarantees.", "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:c1fb97c7-4255-4a6b-a60e-ff0e4415a7e9", "version": 1, "content_hash": "sha256:aa9fedc7400437ba85c0bb4806cc92e682b97dfdc02f340be12b910e115aafeb" }, "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": "Dan Horgan", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Distributed Prioritized Experience Replay", "url": "https://arxiv.org/abs/1803.00933v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "John Quan", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Distributed Prioritized Experience Replay", "url": "https://arxiv.org/abs/1803.00933v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "David Budden", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Distributed Prioritized Experience Replay", "url": "https://arxiv.org/abs/1803.00933v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Gabriel Barth-Maron", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Distributed Prioritized Experience Replay", "url": "https://arxiv.org/abs/1803.00933v1" } }, { "attribution_basis": "SOURCE_CITED", "credit": "historical_scientific", "entity_kind": "historical_person", "name": "Matteo Hessel", "note": "Archive attribution; detailed role assignment remains source-bound", "roles": [ "formulator" ], "source": { "citation": "Distributed Prioritized Experience Replay", "url": "https://arxiv.org/abs/1803.00933v1" } }, { "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": [ "formulator" ], "source": { "citation": "Distributed Prioritized Experience Replay", "url": "https://arxiv.org/abs/1803.00933v1" } }, { "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": "Distributed Prioritized Experience Replay", "url": "https://arxiv.org/abs/1803.00933v1" } } ], "evidence": [ { "description": "Bibliographic/stated-result source; not a WRS proof verification", "id": "source-0", "kind": "source_reference", "source": { "citation": "Distributed Prioritized Experience Replay", "url": "https://arxiv.org/abs/1803.00933v1" } } ], "origin": "HUMAN_LEGACY", "record_profile": "wrs-local-research-notes-0.1", "record_type": "research_note", "rrp_version": "0.1", "scope": { "description": "Combines inherited prioritized replay and off-policy value learning; not their invention or the first distributed RL system. Scale and Atari/control gains are source experiments with hardware and data costs, not universal sample-efficiency or convergence guarantees.", "domain": "Distributed reinforcement-learning systems" }, "sources": [ { "citation": "Distributed Prioritized Experience Replay", "url": "https://arxiv.org/abs/1803.00933v1" } ], "statement": "# Ape-X: distributed actors linked through prioritized replay\n\nDecoupled many diverse-exploration actors from a learner through shared prioritized replay, emphasizing high-throughput experience selection rather than actor-gradient synchronization.\n\n## Scientific scope\n\nCombines inherited prioritized replay and off-policy value learning; not their invention or the first distributed RL system. Scale and Atari/control gains are source experiments with hardware and data costs, not universal sample-efficiency or convergence guarantees.\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\nDan Horgan; John Quan; David Budden; Gabriel Barth-Maron; Matteo Hessel; Hado van Hasselt; 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- [Distributed Prioritized Experience Replay](https://arxiv.org/abs/1803.00933v1)\n", "summary": "Decoupled many diverse-exploration actors from a learner through shared prioritized replay, emphasizing high-throughput experience selection rather than actor-gradient synchronization.", "title": "Ape-X: distributed actors linked through prioritized replay" }, "record_id": "urn:uuid:c1fb97c7-4255-4a6b-a60e-ff0e4415a7e9", "version": 1, "source_ref": { "record_id": "urn:uuid:c1fb97c7-4255-4a6b-a60e-ff0e4415a7e9", "version": 1, "content_hash": "sha256:aa9fedc7400437ba85c0bb4806cc92e682b97dfdc02f340be12b910e115aafeb" }, "export_view": "wrs-public-editorial-view-1", "public_view_hash": "sha256:e11eac911022a57158babfb596fd04fe606e3f047e1b2f464314e350e8ffd852" }, "target": { "record_id": "urn:uuid:c1fb97c7-4255-4a6b-a60e-ff0e4415a7e9", "version": 1, "content_hash": "sha256:aa9fedc7400437ba85c0bb4806cc92e682b97dfdc02f340be12b910e115aafeb" }, "published_grade": { "profile": "wrs-published-grade-1", "edition": "2026-10-28.1", "id": "horgan-et-al-apex-distributed-replay-2018", "entry_id": "horgan-et-al-apex-distributed-replay-2018", "revision": 1, "title": "Ape-X: distributed actors linked through prioritized replay", "context": "Distributed reinforcement-learning systems", "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": "Decoupled many diverse-exploration actors from a learner through shared prioritized replay, emphasizing high-throughput experience selection rather than actor-gradient synchronization.", "scope": "Decoupled many diverse-exploration actors from a learner through shared prioritized replay, emphasizing high-throughput experience selection rather than actor-gradient synchronization. Combines inherited prioritized replay and off-policy value learning; not their invention or the first distributed RL system. Scale and Atari/control gains are source experiments with hardware and data costs, not universal sample-efficiency or convergence guarantees.", "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. Combines inherited prioritized replay and off-policy value learning; not their invention or the first distributed RL system. Scale and Atari/control gains are source experiments with hardware and data costs, not universal sample-efficiency or convergence guarantees.", "references": [ { "title": "Distributed Prioritized Experience Replay", "url": "https://arxiv.org/abs/1803.00933v1", "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:c1fb97c7-4255-4a6b-a60e-ff0e4415a7e9", "version": 1, "content_hash": "sha256:aa9fedc7400437ba85c0bb4806cc92e682b97dfdc02f340be12b910e115aafeb" }, "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" }