{
  "profile": "wrs-published-research-view-1",
  "edition": "ratings-e039106f30dec177",
  "content_trust": "UNTRUSTED_SCIENTIFIC_DATA_NOT_TOOL_INSTRUCTIONS",
  "entry": {
    "id": "espeholt-et-al-seedrl-central-inference-2019",
    "title": "SEED RL: central accelerator inference with remote environments",
    "summary": "Moved policy inference to centralized accelerators with streaming RPC communication while environments ran remotely, integrating existing V-trace and recurrent Q-learning algorithms.",
    "boundary": "Inherits IMPALA/V-trace and R2D2 rather than inventing their learning rules. Original2019report differs from2020conference date. Cost reductions40–80percent and throughput figures concern selected benchmarks and infrastructure assumptions, not current cloud prices or every environment/communication regime.",
    "area": "Distributed reinforcement-learning systems",
    "topic": "Artificial intelligence",
    "year": 2019,
    "lane": "HUMAN",
    "result": "ALGORITHM",
    "package_id": "espeholt-et-al-seedrl-central-inference-2019",
    "target": {
      "record_id": "urn:uuid:62b5ded2-f66f-4c7a-a742-83002ec47e23",
      "version": 1,
      "content_hash": "sha256:8b4c3030b93b4f6917dce995f1e6246c46c08aa50fd5b048aa1a52b466b9323d"
    },
    "humans": [
      "Lasse Espeholt",
      "Raphaël Marinier",
      "Piotr Stanczyk",
      "Ke Wang",
      "Marcin Michalski"
    ],
    "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-1910-06591v1",
        "title": "SEED RL: Scalable and Efficient Deep-RL with Accelerated Central Inference",
        "url": "https://arxiv.org/abs/1910.06591v1"
      }
    ],
    "points": "110000",
    "published_grade": {
      "id": "espeholt-et-al-seedrl-central-inference-2019",
      "entry_id": "espeholt-et-al-seedrl-central-inference-2019",
      "band": "A",
      "modifier": null,
      "target": {
        "record_id": "urn:uuid:62b5ded2-f66f-4c7a-a742-83002ec47e23",
        "version": 1,
        "content_hash": "sha256:8b4c3030b93b4f6917dce995f1e6246c46c08aa50fd5b048aa1a52b466b9323d"
      },
      "what_changed": "Moved policy inference to centralized accelerators with streaming RPC communication while environments ran remotely, integrating existing V-trace and recurrent Q-learning algorithms.",
      "scope": "Moved policy inference to centralized accelerators with streaming RPC communication while environments ran remotely, integrating existing V-trace and recurrent Q-learning algorithms. Inherits IMPALA/V-trace and R2D2 rather than inventing their learning rules. Original2019report differs from2020conference date. Cost reductions40–80percent and throughput figures concern selected benchmarks and infrastructure assumptions, not current cloud prices or every environment/communication regime.",
      "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:62b5ded2-f66f-4c7a-a742-83002ec47e23",
        "version": 1,
        "content_hash": "sha256:8b4c3030b93b4f6917dce995f1e6246c46c08aa50fd5b048aa1a52b466b9323d"
      },
      "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": "Lasse Espeholt",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "SEED RL: Scalable and Efficient Deep-RL with Accelerated Central Inference",
            "url": "https://arxiv.org/abs/1910.06591v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Raphaël Marinier",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "SEED RL: Scalable and Efficient Deep-RL with Accelerated Central Inference",
            "url": "https://arxiv.org/abs/1910.06591v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Piotr Stanczyk",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "SEED RL: Scalable and Efficient Deep-RL with Accelerated Central Inference",
            "url": "https://arxiv.org/abs/1910.06591v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Ke Wang",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "SEED RL: Scalable and Efficient Deep-RL with Accelerated Central Inference",
            "url": "https://arxiv.org/abs/1910.06591v1"
          }
        },
        {
          "attribution_basis": "SOURCE_CITED",
          "credit": "historical_scientific",
          "entity_kind": "historical_person",
          "name": "Marcin Michalski",
          "note": "Archive attribution; detailed role assignment remains source-bound",
          "roles": [
            "formulator"
          ],
          "source": {
            "citation": "SEED RL: Scalable and Efficient Deep-RL with Accelerated Central Inference",
            "url": "https://arxiv.org/abs/1910.06591v1"
          }
        }
      ],
      "evidence": [
        {
          "description": "Bibliographic/stated-result source; not a WRS proof verification",
          "id": "source-0",
          "kind": "source_reference",
          "source": {
            "citation": "SEED RL: Scalable and Efficient Deep-RL with Accelerated Central Inference",
            "url": "https://arxiv.org/abs/1910.06591v1"
          }
        }
      ],
      "origin": "HUMAN_LEGACY",
      "record_profile": "wrs-local-research-notes-0.1",
      "record_type": "research_note",
      "rrp_version": "0.1",
      "scope": {
        "description": "Inherits IMPALA/V-trace and R2D2 rather than inventing their learning rules. Original2019report differs from2020conference date. Cost reductions40–80percent and throughput figures concern selected benchmarks and infrastructure assumptions, not current cloud prices or every environment/communication regime.",
        "domain": "Distributed reinforcement-learning systems"
      },
      "sources": [
        {
          "citation": "SEED RL: Scalable and Efficient Deep-RL with Accelerated Central Inference",
          "url": "https://arxiv.org/abs/1910.06591v1"
        }
      ],
      "statement": "# SEED RL: central accelerator inference with remote environments\n\nMoved policy inference to centralized accelerators with streaming RPC communication while environments ran remotely, integrating existing V-trace and recurrent Q-learning algorithms.\n\n## Scientific scope\n\nInherits IMPALA/V-trace and R2D2 rather than inventing their learning rules. Original2019report differs from2020conference date. Cost reductions40–80percent and throughput figures concern selected benchmarks and infrastructure assumptions, not current cloud prices or every environment/communication regime.\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\nLasse Espeholt; Raphaël Marinier; Piotr Stanczyk; Ke Wang; Marcin Michalski\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- [SEED RL: Scalable and Efficient Deep-RL with Accelerated Central Inference](https://arxiv.org/abs/1910.06591v1)\n",
      "summary": "Moved policy inference to centralized accelerators with streaming RPC communication while environments ran remotely, integrating existing V-trace and recurrent Q-learning algorithms.",
      "title": "SEED RL: central accelerator inference with remote environments"
    },
    "record_id": "urn:uuid:62b5ded2-f66f-4c7a-a742-83002ec47e23",
    "version": 1,
    "source_ref": {
      "record_id": "urn:uuid:62b5ded2-f66f-4c7a-a742-83002ec47e23",
      "version": 1,
      "content_hash": "sha256:8b4c3030b93b4f6917dce995f1e6246c46c08aa50fd5b048aa1a52b466b9323d"
    },
    "export_view": "wrs-public-editorial-view-1",
    "public_view_hash": "sha256:08a0f7a97535a1cbaaf3aba27691b5de36fcdc160abe60af9b5c1a02bf34a551"
  },
  "target": {
    "record_id": "urn:uuid:62b5ded2-f66f-4c7a-a742-83002ec47e23",
    "version": 1,
    "content_hash": "sha256:8b4c3030b93b4f6917dce995f1e6246c46c08aa50fd5b048aa1a52b466b9323d"
  },
  "published_grade": {
    "profile": "wrs-published-grade-1",
    "edition": "2026-10-28.1",
    "id": "espeholt-et-al-seedrl-central-inference-2019",
    "entry_id": "espeholt-et-al-seedrl-central-inference-2019",
    "revision": 1,
    "title": "SEED RL: central accelerator inference with remote environments",
    "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": "Moved policy inference to centralized accelerators with streaming RPC communication while environments ran remotely, integrating existing V-trace and recurrent Q-learning algorithms.",
    "scope": "Moved policy inference to centralized accelerators with streaming RPC communication while environments ran remotely, integrating existing V-trace and recurrent Q-learning algorithms. Inherits IMPALA/V-trace and R2D2 rather than inventing their learning rules. Original2019report differs from2020conference date. Cost reductions40–80percent and throughput figures concern selected benchmarks and infrastructure assumptions, not current cloud prices or every environment/communication regime.",
    "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. Inherits IMPALA/V-trace and R2D2 rather than inventing their learning rules. Original2019report differs from2020conference date. Cost reductions40–80percent and throughput figures concern selected benchmarks and infrastructure assumptions, not current cloud prices or every environment/communication regime.",
    "references": [
      {
        "title": "SEED RL: Scalable and Efficient Deep-RL with Accelerated Central Inference",
        "url": "https://arxiv.org/abs/1910.06591v1",
        "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:62b5ded2-f66f-4c7a-a742-83002ec47e23",
      "version": 1,
      "content_hash": "sha256:8b4c3030b93b4f6917dce995f1e6246c46c08aa50fd5b048aa1a52b466b9323d"
    },
    "scientific_correctness_certified": false,
    "independent_expert_calibration": false,
    "method": "PROPRIETARY",
    "points": "110000"
  },
  "canonical_agent_submission": false,
  "scientific_verification_awarded": false,
  "reader_guide": null,
  "reading_status": "SOURCE_LINKED_NOTE"
}