{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:FDLQ2U6PWJCO6D54DWULJNDTTZ","short_pith_number":"pith:FDLQ2U6P","schema_version":"1.0","canonical_sha256":"28d70d53cfb244ef0fbc1da8b4b4739e7a0ff5d318f50b25384009e733868aa3","source":{"kind":"arxiv","id":"2102.04939","version":1},"attestation_state":"computed","paper":{"title":"RL for Latent MDPs: Regret Guarantees and a Lower Bound","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Constantine Caramanis, Jeongyeol Kwon, Shie Mannor, Yonathan Efroni","submitted_at":"2021-02-09T16:49:58Z","abstract_excerpt":"In this work, we consider the regret minimization problem for reinforcement learning in latent Markov Decision Processes (LMDP). In an LMDP, an MDP is randomly drawn from a set of $M$ possible MDPs at the beginning of the interaction, but the identity of the chosen MDP is not revealed to the agent. We first show that a general instance of LMDPs requires at least $\\Omega((SA)^M)$ episodes to even approximate the optimal policy. Then, we consider sufficient assumptions under which learning good policies requires polynomial number of episodes. We show that the key link is a notion of separation b"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2102.04939","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-02-09T16:49:58Z","cross_cats_sorted":[],"title_canon_sha256":"d171dc5535f7a1d443d5904c95dd8d6d7d74d89f83abba41f0c10385bca5c328","abstract_canon_sha256":"654be6527e1c462576727c8c752a97e44f73801dc06a6ebd7047bff41e9fb422"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:14:06.505801Z","signature_b64":"WLwf1P6LKvOVKC4JhIbZtfySlJvSUqHTq/nb7AApt42i892r7K1gRLrtqBzndbqTx+/y9YOqOix5Zc5VLiGuBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"28d70d53cfb244ef0fbc1da8b4b4739e7a0ff5d318f50b25384009e733868aa3","last_reissued_at":"2026-07-05T02:14:06.505361Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:14:06.505361Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"RL for Latent MDPs: Regret Guarantees and a Lower Bound","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Constantine Caramanis, Jeongyeol Kwon, Shie Mannor, Yonathan Efroni","submitted_at":"2021-02-09T16:49:58Z","abstract_excerpt":"In this work, we consider the regret minimization problem for reinforcement learning in latent Markov Decision Processes (LMDP). In an LMDP, an MDP is randomly drawn from a set of $M$ possible MDPs at the beginning of the interaction, but the identity of the chosen MDP is not revealed to the agent. We first show that a general instance of LMDPs requires at least $\\Omega((SA)^M)$ episodes to even approximate the optimal policy. Then, we consider sufficient assumptions under which learning good policies requires polynomial number of episodes. We show that the key link is a notion of separation b"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2102.04939","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2102.04939/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2102.04939","created_at":"2026-07-05T02:14:06.505424+00:00"},{"alias_kind":"arxiv_version","alias_value":"2102.04939v1","created_at":"2026-07-05T02:14:06.505424+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2102.04939","created_at":"2026-07-05T02:14:06.505424+00:00"},{"alias_kind":"pith_short_12","alias_value":"FDLQ2U6PWJCO","created_at":"2026-07-05T02:14:06.505424+00:00"},{"alias_kind":"pith_short_16","alias_value":"FDLQ2U6PWJCO6D54","created_at":"2026-07-05T02:14:06.505424+00:00"},{"alias_kind":"pith_short_8","alias_value":"FDLQ2U6P","created_at":"2026-07-05T02:14:06.505424+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2502.01800","citing_title":"Flow-based Domain Randomization for Learning and Sequencing Robotic Skills","ref_index":16,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FDLQ2U6PWJCO6D54DWULJNDTTZ","json":"https://pith.science/pith/FDLQ2U6PWJCO6D54DWULJNDTTZ.json","graph_json":"https://pith.science/api/pith-number/FDLQ2U6PWJCO6D54DWULJNDTTZ/graph.json","events_json":"https://pith.science/api/pith-number/FDLQ2U6PWJCO6D54DWULJNDTTZ/events.json","paper":"https://pith.science/paper/FDLQ2U6P"},"agent_actions":{"view_html":"https://pith.science/pith/FDLQ2U6PWJCO6D54DWULJNDTTZ","download_json":"https://pith.science/pith/FDLQ2U6PWJCO6D54DWULJNDTTZ.json","view_paper":"https://pith.science/paper/FDLQ2U6P","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2102.04939&json=true","fetch_graph":"https://pith.science/api/pith-number/FDLQ2U6PWJCO6D54DWULJNDTTZ/graph.json","fetch_events":"https://pith.science/api/pith-number/FDLQ2U6PWJCO6D54DWULJNDTTZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FDLQ2U6PWJCO6D54DWULJNDTTZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FDLQ2U6PWJCO6D54DWULJNDTTZ/action/storage_attestation","attest_author":"https://pith.science/pith/FDLQ2U6PWJCO6D54DWULJNDTTZ/action/author_attestation","sign_citation":"https://pith.science/pith/FDLQ2U6PWJCO6D54DWULJNDTTZ/action/citation_signature","submit_replication":"https://pith.science/pith/FDLQ2U6PWJCO6D54DWULJNDTTZ/action/replication_record"}},"created_at":"2026-07-05T02:14:06.505424+00:00","updated_at":"2026-07-05T02:14:06.505424+00:00"}