{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:FBAO75FPWXQCOAXTYAPINWYMZY","short_pith_number":"pith:FBAO75FP","schema_version":"1.0","canonical_sha256":"2840eff4afb5e02702f3c01e86db0cce10a9aa85f3a5bf7ca9ba9c1db08c9e42","source":{"kind":"arxiv","id":"2303.11369","version":2},"attestation_state":"computed","paper":{"title":"Bridging Imitation and Online Reinforcement Learning: An Optimistic Tale","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Botao Hao, Dengwang Tang, Rahul Jain, Zheng Wen","submitted_at":"2023-03-20T18:16:25Z","abstract_excerpt":"In this paper, we address the following problem: Given an offline demonstration dataset from an imperfect expert, what is the best way to leverage it to bootstrap online learning performance in MDPs. We first propose an Informed Posterior Sampling-based RL (iPSRL) algorithm that uses the offline dataset, and information about the expert's behavioral policy used to generate the offline dataset. Its cumulative Bayesian regret goes down to zero exponentially fast in N, the offline dataset size if the expert is competent enough. Since this algorithm is computationally impractical, we then propose "},"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":"2303.11369","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-03-20T18:16:25Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"6c3ad5fedc9e2f6749cb7b0ba293f082b72bcab756d79bc8a0e20bb35599736e","abstract_canon_sha256":"ae7c61e9ba8508937734127712720902a8e37f079cc22d37948cd2694a8abaed"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:31:10.402561Z","signature_b64":"Z/wM42BTyaSUg1/+cJwEF3VjsuvoZ2mtJ3Q+TiB3g5/jnfRmsGThsMGvhC45VFrA7RU7UUWlLAPoRmNG4iHkAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2840eff4afb5e02702f3c01e86db0cce10a9aa85f3a5bf7ca9ba9c1db08c9e42","last_reissued_at":"2026-07-05T06:31:10.401961Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:31:10.401961Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Bridging Imitation and Online Reinforcement Learning: An Optimistic Tale","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Botao Hao, Dengwang Tang, Rahul Jain, Zheng Wen","submitted_at":"2023-03-20T18:16:25Z","abstract_excerpt":"In this paper, we address the following problem: Given an offline demonstration dataset from an imperfect expert, what is the best way to leverage it to bootstrap online learning performance in MDPs. We first propose an Informed Posterior Sampling-based RL (iPSRL) algorithm that uses the offline dataset, and information about the expert's behavioral policy used to generate the offline dataset. Its cumulative Bayesian regret goes down to zero exponentially fast in N, the offline dataset size if the expert is competent enough. Since this algorithm is computationally impractical, we then propose "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2303.11369","kind":"arxiv","version":2},"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/2303.11369/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":"2303.11369","created_at":"2026-07-05T06:31:10.402027+00:00"},{"alias_kind":"arxiv_version","alias_value":"2303.11369v2","created_at":"2026-07-05T06:31:10.402027+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2303.11369","created_at":"2026-07-05T06:31:10.402027+00:00"},{"alias_kind":"pith_short_12","alias_value":"FBAO75FPWXQC","created_at":"2026-07-05T06:31:10.402027+00:00"},{"alias_kind":"pith_short_16","alias_value":"FBAO75FPWXQCOAXT","created_at":"2026-07-05T06:31:10.402027+00:00"},{"alias_kind":"pith_short_8","alias_value":"FBAO75FP","created_at":"2026-07-05T06:31:10.402027+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2501.18873","citing_title":"Best Policy Learning from Trajectory Preference Feedback","ref_index":2,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FBAO75FPWXQCOAXTYAPINWYMZY","json":"https://pith.science/pith/FBAO75FPWXQCOAXTYAPINWYMZY.json","graph_json":"https://pith.science/api/pith-number/FBAO75FPWXQCOAXTYAPINWYMZY/graph.json","events_json":"https://pith.science/api/pith-number/FBAO75FPWXQCOAXTYAPINWYMZY/events.json","paper":"https://pith.science/paper/FBAO75FP"},"agent_actions":{"view_html":"https://pith.science/pith/FBAO75FPWXQCOAXTYAPINWYMZY","download_json":"https://pith.science/pith/FBAO75FPWXQCOAXTYAPINWYMZY.json","view_paper":"https://pith.science/paper/FBAO75FP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2303.11369&json=true","fetch_graph":"https://pith.science/api/pith-number/FBAO75FPWXQCOAXTYAPINWYMZY/graph.json","fetch_events":"https://pith.science/api/pith-number/FBAO75FPWXQCOAXTYAPINWYMZY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FBAO75FPWXQCOAXTYAPINWYMZY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FBAO75FPWXQCOAXTYAPINWYMZY/action/storage_attestation","attest_author":"https://pith.science/pith/FBAO75FPWXQCOAXTYAPINWYMZY/action/author_attestation","sign_citation":"https://pith.science/pith/FBAO75FPWXQCOAXTYAPINWYMZY/action/citation_signature","submit_replication":"https://pith.science/pith/FBAO75FPWXQCOAXTYAPINWYMZY/action/replication_record"}},"created_at":"2026-07-05T06:31:10.402027+00:00","updated_at":"2026-07-05T06:31:10.402027+00:00"}