{"id":"74f98943234de5d11bf0c8f9ddaeceb6b1448ee84a103787b45d816436c6b232","title":"Finite-Time Convergence and Regret Guarantees for Neural Reinforcement Learning in POMDPs","argument":"Authors Jingjuan Huang, Zhiyao Zhang, Alvaro Velasquez, Nathaniel D. Bastian and Jia Liu argue that their work addresses a gap in finite-time analysis for reinforcement learning in partially observable settings. They say that long-range dependence on history makes POMDPs computationally challenging. Prior analyses of superstate MDPs—approximations built from a finite window of history—either use linear function approximation or rely on an i.i.d. sampling oracle, which the authors say POMDP trajectories violate. The authors analyze nonlinear neural value function approximation, applying Temporal-Difference learning to superstate transitions and combining a critic error bound with policy mirror descent. They report finite-time bounds and sublinear regret against the optimal POMDP value, and describe this as the first such result based on neural function approximation for finite-memory POMDPs. Escape Meridian interpretation: if the analysis holds under its stated conditions, it offers a theoretical route beyond the cited simplifications; it does not by itself establish performance in practical deployments. The authors also say numerical experiments verify their theoretical results, but the abstract does not describe those experiments.","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The abstract does not describe the numerical experiments, their setup, or their results. The abstract reports theoretical bounds but does not provide their precise conditions or quantities.","speaker":{"name":"Huang, Jingjuan, Zhang, Zhiyao, Velasquez, Alvaro, Bastian, Nathaniel D., Liu, Jia","roleAtSource":"Article author"},"source":{"title":"Finite-Time Convergence and Regret Guarantees for Neural Reinforcement Learning in POMDPs","publisher":"arXiv cs.AI","format":"Article","url":"https://arxiv.org/abs/2611.07069","publishedAt":"2026-11-04","dateBasis":"Date basis not provided."},"evidence":[{"text":"The authors say existing finite-time analyses of superstate MDPs are limited by linear function approximation or reliance on an i.i.d. sampling oracle that POMDP trajectories violate.","excerpt":"either limited to over-simplified linear function approximation or rely on an i.i.d. sampling oracle that is violated by POMDP trajectories","url":"https://arxiv.org/abs/2611.07069","relationship":"Source-stated"},{"text":"The authors report a finite-time critic error bound using nonlinear neural value function approximation and a finite-time sublinear regret bound against the optimal POMDP value.","excerpt":"we obtain a finite-time sublinear regret against the optimal POMDP value","url":"https://arxiv.org/abs/2611.07069","relationship":"Source-stated"},{"text":"The authors say their numerical experiments verify their theoretical results.","excerpt":"Our numerical experiments also verify our theoretical results.","url":"https://arxiv.org/abs/2611.07069","relationship":"Source-stated"},{"text":"If the analysis holds under its stated conditions, it offers a theoretical route beyond the cited simplifications, but does not by itself establish performance in practical deployments.","excerpt":"both of which severely limit the applicability of these theoretical results in practice","url":"https://arxiv.org/abs/2611.07069","relationship":"Escape Meridian interpretation"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2611.07069","checkedAt":"2026-11-09T00:34:07.710Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"What numerical experiments did the authors run, and how do their results support the theoretical claims?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"Finite-Time Convergence and Regret Guarantees for Neural Reinforcement Learning in POMDPs","summary":"Authors Jingjuan Huang, Zhiyao Zhang, Alvaro Velasquez, Nathaniel D. Bastian and Jia Liu argue that their work addresses a gap in finite-time analysis for reinforcement learning in partially observable settings. They say that long-range dependence on history makes POMDPs computationally challenging. Prior analyses of superstate MDPs—approximations built from a finite window of history—either use linear function approximation or rely on an i.i.d. sampling oracle, which the authors say POMDP trajectories violate. The authors analyze nonlinear neural value function approximation, applying Temporal-Difference learning to superstate transitions and combining a critic error bound with policy mirror descent. They report finite-time bounds and sublinear regret against the optimal POMDP value, and describe this as the first such result based on neural function approximation for finite-memory POMDPs. Escape Meridian interpretation: if the analysis holds under its stated conditions, it offers a theoretical route beyond the cited simplifications; it does not by itself establish performance in practical deployments. The authors also say numerical experiments verify their theoretical results, but the abstract does not describe those experiments.","question":"What numerical experiments did the authors run, and how do their results support the theoretical claims?","claims":[{"text":"The authors say existing finite-time analyses of superstate MDPs are limited by linear function approximation or reliance on an i.i.d. sampling oracle that POMDP trajectories violate.","excerpt":"either limited to over-simplified linear function approximation or rely on an i.i.d. sampling oracle that is violated by POMDP trajectories","support":"source-stated"},{"text":"The authors report a finite-time critic error bound using nonlinear neural value function approximation and a finite-time sublinear regret bound against the optimal POMDP value.","excerpt":"we obtain a finite-time sublinear regret against the optimal POMDP value","support":"source-stated"},{"text":"The authors say their numerical experiments verify their theoretical results.","excerpt":"Our numerical experiments also verify our theoretical results.","support":"source-stated"},{"text":"If the analysis holds under its stated conditions, it offers a theoretical route beyond the cited simplifications, but does not by itself establish performance in practical deployments.","excerpt":"both of which severely limit the applicability of these theoretical results in practice","support":"interpretation"}],"limitations":["The abstract does not describe the numerical experiments, their setup, or their results.","The abstract reports theoretical bounds but does not provide their precise conditions or quantities."],"mediaId":"b725db599f9d0347d9dee481b3a1312c7be1c0d3b0db3637236914bf9302eab7","correctionReason":null,"id":"74f98943234de5d11bf0c8f9ddaeceb6b1448ee84a103787b45d816436c6b232","publicationId":"d68f0b97-70ba-4620-8b42-e290a429ec27","version":1,"sourceUrl":"https://arxiv.org/abs/2611.07069","sourceTitle":"Finite-Time Convergence and Regret Guarantees for Neural Reinforcement Learning in POMDPs","publisher":"arXiv cs.AI","publishedAt":null,"originalPublished":{"value":"2026-11-04","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-11-06T05:10:19.119Z","createdAt":"2026-11-09T00:02:50.345Z","historical":true,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"b725db599f9d0347d9dee481b3a1312c7be1c0d3b0db3637236914bf9302eab7","kind":"article","url":"https://arxiv.org/abs/2611.07069","sourceUrl":"https://arxiv.org/abs/2611.07069","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-11-09T00:01:58.054Z","method":"acquired_source","contentType":"text/html"},"recommendationReason":"Recommended by Jev for understanding this source; checked by Luna."}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2611.07069","checkedAt":"2026-11-09T00:34:07.710Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["material_development"],"rubricVersion":"stream-admission-v5","mediaId":"b725db599f9d0347d9dee481b3a1312c7be1c0d3b0db3637236914bf9302eab7","confidence":0.64,"provider":"typesafe","model":"jev-1.13.0","inputHash":"9168ef0571a5d8834fbbe8b4f318ed95f973e2d5ce1225a569d733f057f4f411","decidedAt":"2026-11-09T00:01:58.831Z"},"provenance":{"textHash":"f29045731b1ca66a896136b0115c61d083d2683d7fbfcf24519e0810dcd5d6b0","observationId":"9f570105b64e3cf8554b1509f340afa421fb9a23ea74859712458f0b599398fa","retrievedAt":"2026-11-09T00:01:57.939Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"aee98bd7-e030-45cb-87ae-6a4137c4fa69"},"reviewer":{"model":"gpt-6-luna","runId":"697a2141-1e55-49c0-88e5-d31c131504c5"}},"audit":{"status":"uncertain","reviewedAt":"2026-11-09T00:02:50.345Z","model":"anthropic/claude-opus-5.5"},"sourceAttribution":{"authors":[{"name":"Huang, Jingjuan","url":null,"kind":"unknown"},{"name":"Zhang, Zhiyao","url":null,"kind":"unknown"},{"name":"Velasquez, Alvaro","url":null,"kind":"unknown"},{"name":"Bastian, Nathaniel D.","url":null,"kind":"unknown"},{"name":"Liu, Jia","url":null,"kind":"unknown"}],"authorRole":"article-author","hosts":[],"status":"stated","authorRoster":{"truncated":false,"omitted":0,"assertionGroupsOmitted":0},"participation":[{"actor":{"name":"Huang, Jingjuan","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Huang, Jingjuan","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.07069","locator":"head/meta[name=author|citation_author]"},"assertionId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97:participation:0","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T00:34:07.710Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97","sourceUrl":"https://arxiv.org/abs/2611.07069","contentSha256":"e39dc565da6d2b1aa4297be6a176c2e9f51931566fda00f63a4c2414efe085f9","textHash":"f29045731b1ca66a896136b0115c61d083d2683d7fbfcf24519e0810dcd5d6b0","observedAt":"2026-11-09T00:34:07.710Z"}},{"actor":{"name":"Zhang, Zhiyao","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Zhang, Zhiyao","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.07069","locator":"head/meta[name=author|citation_author]"},"assertionId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97:participation:1","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T00:34:07.710Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97","sourceUrl":"https://arxiv.org/abs/2611.07069","contentSha256":"e39dc565da6d2b1aa4297be6a176c2e9f51931566fda00f63a4c2414efe085f9","textHash":"f29045731b1ca66a896136b0115c61d083d2683d7fbfcf24519e0810dcd5d6b0","observedAt":"2026-11-09T00:34:07.710Z"}},{"actor":{"name":"Velasquez, Alvaro","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Velasquez, Alvaro","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.07069","locator":"head/meta[name=author|citation_author]"},"assertionId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97:participation:2","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T00:34:07.710Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97","sourceUrl":"https://arxiv.org/abs/2611.07069","contentSha256":"e39dc565da6d2b1aa4297be6a176c2e9f51931566fda00f63a4c2414efe085f9","textHash":"f29045731b1ca66a896136b0115c61d083d2683d7fbfcf24519e0810dcd5d6b0","observedAt":"2026-11-09T00:34:07.710Z"}},{"actor":{"name":"Bastian, Nathaniel D.","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Bastian, Nathaniel D.","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.07069","locator":"head/meta[name=author|citation_author]"},"assertionId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97:participation:3","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T00:34:07.710Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97","sourceUrl":"https://arxiv.org/abs/2611.07069","contentSha256":"e39dc565da6d2b1aa4297be6a176c2e9f51931566fda00f63a4c2414efe085f9","textHash":"f29045731b1ca66a896136b0115c61d083d2683d7fbfcf24519e0810dcd5d6b0","observedAt":"2026-11-09T00:34:07.710Z"}},{"actor":{"name":"Liu, Jia","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Liu, Jia","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.07069","locator":"head/meta[name=author|citation_author]"},"assertionId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97:participation:4","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T00:34:07.710Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97","sourceUrl":"https://arxiv.org/abs/2611.07069","contentSha256":"e39dc565da6d2b1aa4297be6a176c2e9f51931566fda00f63a4c2414efe085f9","textHash":"f29045731b1ca66a896136b0115c61d083d2683d7fbfcf24519e0810dcd5d6b0","observedAt":"2026-11-09T00:34:07.710Z"}}],"participationCoverage":{"total":5,"returned":5,"complete":true,"fullAssertions":"source-metadata-endpoint"},"distribution":null,"publisher":null,"sourceUrl":"https://arxiv.org/abs/2611.07069","snapshotId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97","contentSha256":"e39dc565da6d2b1aa4297be6a176c2e9f51931566fda00f63a4c2414efe085f9","textHash":"f29045731b1ca66a896136b0115c61d083d2683d7fbfcf24519e0810dcd5d6b0","observedAt":"2026-11-09T00:34:07.710Z","basis":"publisher-page-metadata","sourceLinks":45,"linksTruncated":false,"metadataUrl":"/api/pipeline/stream/74f98943234de5d11bf0c8f9ddaeceb6b1448ee84a103787b45d816436c6b232/source"}},"canonicalUrl":"https://escapemeridian.com/readings/finite-time-convergence-and-regret-guarantees-for-neural-74f98943","formats":{"html":"/readings/finite-time-convergence-and-regret-guarantees-for-neural-74f98943","markdown":"/readings/finite-time-convergence-and-regret-guarantees-for-neural-74f98943.md","json":"/readings/finite-time-convergence-and-regret-guarantees-for-neural-74f98943.json"},"attribution":"Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.","maintainedGuideLinks":[],"readerContext":{"version":"reader-context-v1","attribution":"Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.","terms":[{"id":"term-reinforcement-learning","name":"Reinforcement learning","explanation":"Training by trial and error, in which a system is rewarded for the outcomes its designers want.","referenceUrl":"https://developers.google.com/machine-learning/glossary#reinforcement-learning-rl"}],"entities":[{"id":"time","name":"TIME","kind":"organization","context":"TIME is a news magazine publishing reporting on world affairs, business, technology and the workers behind digital services.","referenceUrls":["https://time.com/"],"path":"/entities/time","json":"/entities/time.json","markdown":"/entities/time.md"},{"id":"arxiv","name":"arXiv","kind":"organization","context":"arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.","referenceUrls":["https://arxiv.org/","https://info.arxiv.org/help/api/index.html"],"path":"/entities/arxiv","json":"/entities/arxiv.json","markdown":"/entities/arxiv.md"}]},"researchQuestions":[],"withdrawn":false,"references":{"schemaVersion":1,"target":{"type":"reading","id":"74f98943234de5d11bf0c8f9ddaeceb6b1448ee84a103787b45d816436c6b232","revision":"d68f0b97-70ba-4620-8b42-e290a429ec27","revisionKind":"publication-id","title":"Finite-Time Convergence and Regret Guarantees for Neural Reinforcement Learning in POMDPs","url":"https://escapemeridian.com/readings/finite-time-convergence-and-regret-guarantees-for-neural-74f98943"},"source":{"type":"source","id":"https://arxiv.org/abs/2611.07069","revision":"f29045731b1ca66a896136b0115c61d083d2683d7fbfcf24519e0810dcd5d6b0","revisionKind":"sha256-source-text","title":"Finite-Time Convergence and Regret Guarantees for Neural Reinforcement Learning in POMDPs","url":"https://arxiv.org/abs/2611.07069"},"status":"available","items":[],"next":null,"scope":"Explicit citations in public readings and public file attachments to this reading. Shared topics are not citations. Other artifact types need an explicit reference adapter."},"revisions":[{"id":"74f98943234de5d11bf0c8f9ddaeceb6b1448ee84a103787b45d816436c6b232","title":"Finite-Time Convergence and Regret Guarantees for Neural Reinforcement Learning in POMDPs","argument":"Authors Jingjuan Huang, Zhiyao Zhang, Alvaro Velasquez, Nathaniel D. Bastian and Jia Liu argue that their work addresses a gap in finite-time analysis for reinforcement learning in partially observable settings. They say that long-range dependence on history makes POMDPs computationally challenging. Prior analyses of superstate MDPs—approximations built from a finite window of history—either use linear function approximation or rely on an i.i.d. sampling oracle, which the authors say POMDP trajectories violate. The authors analyze nonlinear neural value function approximation, applying Temporal-Difference learning to superstate transitions and combining a critic error bound with policy mirror descent. They report finite-time bounds and sublinear regret against the optimal POMDP value, and describe this as the first such result based on neural function approximation for finite-memory POMDPs. Escape Meridian interpretation: if the analysis holds under its stated conditions, it offers a theoretical route beyond the cited simplifications; it does not by itself establish performance in practical deployments. The authors also say numerical experiments verify their theoretical results, but the abstract does not describe those experiments.","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The abstract does not describe the numerical experiments, their setup, or their results. The abstract reports theoretical bounds but does not provide their precise conditions or quantities.","speaker":{"name":"Huang, Jingjuan, Zhang, Zhiyao, Velasquez, Alvaro, Bastian, Nathaniel D., Liu, Jia","roleAtSource":"Article author"},"source":{"title":"Finite-Time Convergence and Regret Guarantees for Neural Reinforcement Learning in POMDPs","publisher":"arXiv cs.AI","format":"Article","url":"https://arxiv.org/abs/2611.07069","publishedAt":"2026-11-04","dateBasis":"Date basis not provided."},"evidence":[{"text":"The authors say existing finite-time analyses of superstate MDPs are limited by linear function approximation or reliance on an i.i.d. sampling oracle that POMDP trajectories violate.","excerpt":"either limited to over-simplified linear function approximation or rely on an i.i.d. sampling oracle that is violated by POMDP trajectories","url":"https://arxiv.org/abs/2611.07069","relationship":"Source-stated"},{"text":"The authors report a finite-time critic error bound using nonlinear neural value function approximation and a finite-time sublinear regret bound against the optimal POMDP value.","excerpt":"we obtain a finite-time sublinear regret against the optimal POMDP value","url":"https://arxiv.org/abs/2611.07069","relationship":"Source-stated"},{"text":"The authors say their numerical experiments verify their theoretical results.","excerpt":"Our numerical experiments also verify our theoretical results.","url":"https://arxiv.org/abs/2611.07069","relationship":"Source-stated"},{"text":"If the analysis holds under its stated conditions, it offers a theoretical route beyond the cited simplifications, but does not by itself establish performance in practical deployments.","excerpt":"both of which severely limit the applicability of these theoretical results in practice","url":"https://arxiv.org/abs/2611.07069","relationship":"Escape Meridian interpretation"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2611.07069","checkedAt":"2026-11-09T00:34:07.710Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"What numerical experiments did the authors run, and how do their results support the theoretical claims?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"Finite-Time Convergence and Regret Guarantees for Neural Reinforcement Learning in POMDPs","summary":"Authors Jingjuan Huang, Zhiyao Zhang, Alvaro Velasquez, Nathaniel D. Bastian and Jia Liu argue that their work addresses a gap in finite-time analysis for reinforcement learning in partially observable settings. They say that long-range dependence on history makes POMDPs computationally challenging. Prior analyses of superstate MDPs—approximations built from a finite window of history—either use linear function approximation or rely on an i.i.d. sampling oracle, which the authors say POMDP trajectories violate. The authors analyze nonlinear neural value function approximation, applying Temporal-Difference learning to superstate transitions and combining a critic error bound with policy mirror descent. They report finite-time bounds and sublinear regret against the optimal POMDP value, and describe this as the first such result based on neural function approximation for finite-memory POMDPs. Escape Meridian interpretation: if the analysis holds under its stated conditions, it offers a theoretical route beyond the cited simplifications; it does not by itself establish performance in practical deployments. The authors also say numerical experiments verify their theoretical results, but the abstract does not describe those experiments.","question":"What numerical experiments did the authors run, and how do their results support the theoretical claims?","claims":[{"text":"The authors say existing finite-time analyses of superstate MDPs are limited by linear function approximation or reliance on an i.i.d. sampling oracle that POMDP trajectories violate.","excerpt":"either limited to over-simplified linear function approximation or rely on an i.i.d. sampling oracle that is violated by POMDP trajectories","support":"source-stated"},{"text":"The authors report a finite-time critic error bound using nonlinear neural value function approximation and a finite-time sublinear regret bound against the optimal POMDP value.","excerpt":"we obtain a finite-time sublinear regret against the optimal POMDP value","support":"source-stated"},{"text":"The authors say their numerical experiments verify their theoretical results.","excerpt":"Our numerical experiments also verify our theoretical results.","support":"source-stated"},{"text":"If the analysis holds under its stated conditions, it offers a theoretical route beyond the cited simplifications, but does not by itself establish performance in practical deployments.","excerpt":"both of which severely limit the applicability of these theoretical results in practice","support":"interpretation"}],"limitations":["The abstract does not describe the numerical experiments, their setup, or their results.","The abstract reports theoretical bounds but does not provide their precise conditions or quantities."],"mediaId":"b725db599f9d0347d9dee481b3a1312c7be1c0d3b0db3637236914bf9302eab7","correctionReason":null,"id":"74f98943234de5d11bf0c8f9ddaeceb6b1448ee84a103787b45d816436c6b232","publicationId":"d68f0b97-70ba-4620-8b42-e290a429ec27","version":1,"sourceUrl":"https://arxiv.org/abs/2611.07069","sourceTitle":"Finite-Time Convergence and Regret Guarantees for Neural Reinforcement Learning in POMDPs","publisher":"arXiv cs.AI","publishedAt":null,"originalPublished":{"value":"2026-11-04","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-11-06T05:10:19.119Z","createdAt":"2026-11-09T00:02:50.345Z","historical":true,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"b725db599f9d0347d9dee481b3a1312c7be1c0d3b0db3637236914bf9302eab7","kind":"article","url":"https://arxiv.org/abs/2611.07069","sourceUrl":"https://arxiv.org/abs/2611.07069","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-11-09T00:01:58.054Z","method":"acquired_source","contentType":"text/html"},"recommendationReason":"Recommended by Jev for understanding this source; checked by Luna."}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2611.07069","checkedAt":"2026-11-09T00:34:07.710Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["material_development"],"rubricVersion":"stream-admission-v5","mediaId":"b725db599f9d0347d9dee481b3a1312c7be1c0d3b0db3637236914bf9302eab7","confidence":0.64,"provider":"typesafe","model":"jev-1.13.0","inputHash":"9168ef0571a5d8834fbbe8b4f318ed95f973e2d5ce1225a569d733f057f4f411","decidedAt":"2026-11-09T00:01:58.831Z"},"provenance":{"textHash":"f29045731b1ca66a896136b0115c61d083d2683d7fbfcf24519e0810dcd5d6b0","observationId":"9f570105b64e3cf8554b1509f340afa421fb9a23ea74859712458f0b599398fa","retrievedAt":"2026-11-09T00:01:57.939Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"aee98bd7-e030-45cb-87ae-6a4137c4fa69"},"reviewer":{"model":"gpt-6-luna","runId":"697a2141-1e55-49c0-88e5-d31c131504c5"}},"audit":{"status":"uncertain","reviewedAt":"2026-11-09T00:02:50.345Z","model":"anthropic/claude-opus-5.5"},"sourceAttribution":{"authors":[{"name":"Huang, Jingjuan","url":null,"kind":"unknown"},{"name":"Zhang, Zhiyao","url":null,"kind":"unknown"},{"name":"Velasquez, Alvaro","url":null,"kind":"unknown"},{"name":"Bastian, Nathaniel D.","url":null,"kind":"unknown"},{"name":"Liu, Jia","url":null,"kind":"unknown"}],"authorRole":"article-author","hosts":[],"status":"stated","authorRoster":{"truncated":false,"omitted":0,"assertionGroupsOmitted":0},"participation":[{"actor":{"name":"Huang, Jingjuan","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Huang, Jingjuan","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.07069","locator":"head/meta[name=author|citation_author]"},"assertionId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97:participation:0","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T00:34:07.710Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97","sourceUrl":"https://arxiv.org/abs/2611.07069","contentSha256":"e39dc565da6d2b1aa4297be6a176c2e9f51931566fda00f63a4c2414efe085f9","textHash":"f29045731b1ca66a896136b0115c61d083d2683d7fbfcf24519e0810dcd5d6b0","observedAt":"2026-11-09T00:34:07.710Z"}},{"actor":{"name":"Zhang, Zhiyao","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Zhang, Zhiyao","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.07069","locator":"head/meta[name=author|citation_author]"},"assertionId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97:participation:1","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T00:34:07.710Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97","sourceUrl":"https://arxiv.org/abs/2611.07069","contentSha256":"e39dc565da6d2b1aa4297be6a176c2e9f51931566fda00f63a4c2414efe085f9","textHash":"f29045731b1ca66a896136b0115c61d083d2683d7fbfcf24519e0810dcd5d6b0","observedAt":"2026-11-09T00:34:07.710Z"}},{"actor":{"name":"Velasquez, Alvaro","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Velasquez, Alvaro","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.07069","locator":"head/meta[name=author|citation_author]"},"assertionId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97:participation:2","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T00:34:07.710Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97","sourceUrl":"https://arxiv.org/abs/2611.07069","contentSha256":"e39dc565da6d2b1aa4297be6a176c2e9f51931566fda00f63a4c2414efe085f9","textHash":"f29045731b1ca66a896136b0115c61d083d2683d7fbfcf24519e0810dcd5d6b0","observedAt":"2026-11-09T00:34:07.710Z"}},{"actor":{"name":"Bastian, Nathaniel D.","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Bastian, Nathaniel D.","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.07069","locator":"head/meta[name=author|citation_author]"},"assertionId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97:participation:3","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T00:34:07.710Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97","sourceUrl":"https://arxiv.org/abs/2611.07069","contentSha256":"e39dc565da6d2b1aa4297be6a176c2e9f51931566fda00f63a4c2414efe085f9","textHash":"f29045731b1ca66a896136b0115c61d083d2683d7fbfcf24519e0810dcd5d6b0","observedAt":"2026-11-09T00:34:07.710Z"}},{"actor":{"name":"Liu, Jia","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Liu, Jia","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2611.07069","locator":"head/meta[name=author|citation_author]"},"assertionId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97:participation:4","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-09T00:34:07.710Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97","sourceUrl":"https://arxiv.org/abs/2611.07069","contentSha256":"e39dc565da6d2b1aa4297be6a176c2e9f51931566fda00f63a4c2414efe085f9","textHash":"f29045731b1ca66a896136b0115c61d083d2683d7fbfcf24519e0810dcd5d6b0","observedAt":"2026-11-09T00:34:07.710Z"}}],"participationCoverage":{"total":5,"returned":5,"complete":true,"fullAssertions":"source-metadata-endpoint"},"distribution":null,"publisher":null,"sourceUrl":"https://arxiv.org/abs/2611.07069","snapshotId":"f0963ae45d1707fe7e6016ef2d28b682676b862c8ab1581385da0171129d6e97","contentSha256":"e39dc565da6d2b1aa4297be6a176c2e9f51931566fda00f63a4c2414efe085f9","textHash":"f29045731b1ca66a896136b0115c61d083d2683d7fbfcf24519e0810dcd5d6b0","observedAt":"2026-11-09T00:34:07.710Z","basis":"publisher-page-metadata","sourceLinks":45,"linksTruncated":false,"metadataUrl":"/api/pipeline/stream/74f98943234de5d11bf0c8f9ddaeceb6b1448ee84a103787b45d816436c6b232/source"}}}],"audits":[{"publicationId":"d68f0b97-70ba-4620-8b42-e290a429ec27","verdict":"uncertain","findings":[{"dimension":"attribution","status":"pass","explanation":"The five authors match the arXiv listing. The reading attributes the gap, the method and the results to the authors rather than presenting them as established facts.","excerpt":"Authors:Jingjuan Huang, Zhiyao Zhang, Alvaro Velasquez, Nathaniel D. Bastian, Jia Liu"},{"dimension":"summary","status":"uncertain","explanation":"The reading is faithful but not readable for a general audience. Central terms are left unexplained: POMDP (only glossed as partially observable), finite-time bounds, regret and sublinear regret, i.i.d. sampling oracle, Temporal-Difference learning, critic and policy mirror descent. A non-specialist cannot tell what was proven or why it matters. Two small fidelity points: the summary omits the filter stability assumption that underlies the superstate approximation, and it drops the authors' hedge 'to the best of our knowledge' from the first-result claim. Repair: add plain-language glosses or tooltips, and name the filter stability condition.","excerpt":"a range of POMDPs that satisfy the filter stability assumption can be well approximated by a superstate Markov Decision Process (MDP)"},{"dimension":"date","status":"pass","explanation":"originalPublished is 2026-11-04, which matches the v1 submission date. The feed date of 2026-11-06 is correctly not used.","excerpt":"[Submitted on 4 Nov 2026]"},{"dimension":"support","status":"pass","explanation":"Each claim is grounded in the abstract. The interpretation is labeled as Escape Meridian's, stated conditionally, and does not claim any practical benefit. The question follows from the fact that the abstract does not describe the experiments.","excerpt":"we obtain a finite-time sublinear regret against the optimal POMDP value"},{"dimension":"media","status":"pass","explanation":"The selected media is the verified direct link to the arXiv abstract page, which is the original source.","excerpt":null}],"createdAt":"2026-11-09T00:02:50.345Z","model":"anthropic/claude-opus-5.5"}],"correctionUrl":"https://escapemeridian.com/corrections/"}