{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:4APG6YXTEBEVWHLPMPHAFU72HQ","short_pith_number":"pith:4APG6YXT","schema_version":"1.0","canonical_sha256":"e01e6f62f320495b1d6f63ce02d3fa3c02fb2f79468c5dcd500273b9e3a4415b","source":{"kind":"arxiv","id":"2003.02189","version":1},"attestation_state":"computed","paper":{"title":"Exploration-Exploitation in Constrained MDPs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Matteo Pirotta, Shie Mannor, Yonathan Efroni","submitted_at":"2020-03-04T17:03:56Z","abstract_excerpt":"In many sequential decision-making problems, the goal is to optimize a utility function while satisfying a set of constraints on different utilities. This learning problem is formalized through Constrained Markov Decision Processes (CMDPs). In this paper, we investigate the exploration-exploitation dilemma in CMDPs. While learning in an unknown CMDP, an agent should trade-off exploration to discover new information about the MDP, and exploitation of the current knowledge to maximize the reward while satisfying the constraints. While the agent will eventually learn a good or optimal policy, we "},"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":"2003.02189","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-03-04T17:03:56Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"5957b88ef595cb4fd396919fd67577e8582e66229e09c23c3da6228c2fbf62a5","abstract_canon_sha256":"581a9385a13e507ff9ec1061d8b1d610407edbbb966ff33c5be2700fc743a1a0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:45:47.746751Z","signature_b64":"fLyrbELvVEjaebML8HbGi3CY8zP6V4OewGZaROqWDf9YeJ69cQ1DUBdf9HrQqmSolAPSUhpd1Sy/cKUTY7lNCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e01e6f62f320495b1d6f63ce02d3fa3c02fb2f79468c5dcd500273b9e3a4415b","last_reissued_at":"2026-07-05T00:45:47.746255Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:45:47.746255Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Exploration-Exploitation in Constrained MDPs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Matteo Pirotta, Shie Mannor, Yonathan Efroni","submitted_at":"2020-03-04T17:03:56Z","abstract_excerpt":"In many sequential decision-making problems, the goal is to optimize a utility function while satisfying a set of constraints on different utilities. This learning problem is formalized through Constrained Markov Decision Processes (CMDPs). In this paper, we investigate the exploration-exploitation dilemma in CMDPs. While learning in an unknown CMDP, an agent should trade-off exploration to discover new information about the MDP, and exploitation of the current knowledge to maximize the reward while satisfying the constraints. While the agent will eventually learn a good or optimal policy, we "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2003.02189","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/2003.02189/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":"2003.02189","created_at":"2026-07-05T00:45:47.746326+00:00"},{"alias_kind":"arxiv_version","alias_value":"2003.02189v1","created_at":"2026-07-05T00:45:47.746326+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2003.02189","created_at":"2026-07-05T00:45:47.746326+00:00"},{"alias_kind":"pith_short_12","alias_value":"4APG6YXTEBEV","created_at":"2026-07-05T00:45:47.746326+00:00"},{"alias_kind":"pith_short_16","alias_value":"4APG6YXTEBEVWHLP","created_at":"2026-07-05T00:45:47.746326+00:00"},{"alias_kind":"pith_short_8","alias_value":"4APG6YXT","created_at":"2026-07-05T00:45:47.746326+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":9,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.01159","citing_title":"Fairness in two-player zero-sum games with bandit feedback","ref_index":43,"is_internal_anchor":false},{"citing_arxiv_id":"2603.22348","citing_title":"Learning Safely Without Knowing the World:COMPASS-Hedge","ref_index":22,"is_internal_anchor":false},{"citing_arxiv_id":"2605.23351","citing_title":"Prudent-Banker: No Extra Fees for Baseline Safety in Adversarial Bandits With and Without Delays","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2408.16286","citing_title":"Near-Optimal Policy Identification in Robust Constrained Markov Decision Processes via Epigraph Form","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2408.11513","citing_title":"Last-Iterate Convergence of General Parameterized Policies in Constrained MDPs","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11535","citing_title":"Primal-Dual Policy Optimization for Linear CMDPs with Adversarial Losses","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09214","citing_title":"Fast Rates for Offline Contextual Bandits with Forward-KL Regularization under Single-Policy Concentrability","ref_index":59,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10519","citing_title":"Online Resource Allocation With General Constraints","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08290","citing_title":"Toward Optimal Regret in Robust Pricing: Decoupling Corruption and Time","ref_index":13,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4APG6YXTEBEVWHLPMPHAFU72HQ","json":"https://pith.science/pith/4APG6YXTEBEVWHLPMPHAFU72HQ.json","graph_json":"https://pith.science/api/pith-number/4APG6YXTEBEVWHLPMPHAFU72HQ/graph.json","events_json":"https://pith.science/api/pith-number/4APG6YXTEBEVWHLPMPHAFU72HQ/events.json","paper":"https://pith.science/paper/4APG6YXT"},"agent_actions":{"view_html":"https://pith.science/pith/4APG6YXTEBEVWHLPMPHAFU72HQ","download_json":"https://pith.science/pith/4APG6YXTEBEVWHLPMPHAFU72HQ.json","view_paper":"https://pith.science/paper/4APG6YXT","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2003.02189&json=true","fetch_graph":"https://pith.science/api/pith-number/4APG6YXTEBEVWHLPMPHAFU72HQ/graph.json","fetch_events":"https://pith.science/api/pith-number/4APG6YXTEBEVWHLPMPHAFU72HQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4APG6YXTEBEVWHLPMPHAFU72HQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4APG6YXTEBEVWHLPMPHAFU72HQ/action/storage_attestation","attest_author":"https://pith.science/pith/4APG6YXTEBEVWHLPMPHAFU72HQ/action/author_attestation","sign_citation":"https://pith.science/pith/4APG6YXTEBEVWHLPMPHAFU72HQ/action/citation_signature","submit_replication":"https://pith.science/pith/4APG6YXTEBEVWHLPMPHAFU72HQ/action/replication_record"}},"created_at":"2026-07-05T00:45:47.746326+00:00","updated_at":"2026-07-05T00:45:47.746326+00:00"}