{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:TNJKNAHQXG37B7X5S7STNVDBEQ","short_pith_number":"pith:TNJKNAHQ","schema_version":"1.0","canonical_sha256":"9b52a680f0b9b7f0fefd97e536d461241c84e9c9e33d8994924a6b628d5748ae","source":{"kind":"arxiv","id":"2407.04451","version":1},"attestation_state":"computed","paper":{"title":"Hindsight Preference Learning for Offline Preference-based Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Chenjun Xiao, Chen-Xiao Gao, Shengjun Fang, Yang Yu, Zongzhang Zhang","submitted_at":"2024-07-05T12:05:37Z","abstract_excerpt":"Offline preference-based reinforcement learning (RL), which focuses on optimizing policies using human preferences between pairs of trajectory segments selected from an offline dataset, has emerged as a practical avenue for RL applications. Existing works rely on extracting step-wise reward signals from trajectory-wise preference annotations, assuming that preferences correlate with the cumulative Markovian rewards. However, such methods fail to capture the holistic perspective of data annotation: Humans often assess the desirability of a sequence of actions by considering the overall outcome "},"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":"2407.04451","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-07-05T12:05:37Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"03197d4a277593b3da9aa851bfa21d678a52cc7d9c8bf34f2cc32a7b257a04a5","abstract_canon_sha256":"7e63c4fe31cf127041c201b039151fad14941e0e0fdb851d757d968539e2ae79"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:40:33.847666Z","signature_b64":"X6Nv6LGG+tJWodDp5nIcjgsBZxaBWT2KOu7FtcFg3cH45t+FRRcKDj5HzaOae63qmg/S7+6I0wNX4ge5SQg5AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9b52a680f0b9b7f0fefd97e536d461241c84e9c9e33d8994924a6b628d5748ae","last_reissued_at":"2026-07-05T08:40:33.847163Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:40:33.847163Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Hindsight Preference Learning for Offline Preference-based Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Chenjun Xiao, Chen-Xiao Gao, Shengjun Fang, Yang Yu, Zongzhang Zhang","submitted_at":"2024-07-05T12:05:37Z","abstract_excerpt":"Offline preference-based reinforcement learning (RL), which focuses on optimizing policies using human preferences between pairs of trajectory segments selected from an offline dataset, has emerged as a practical avenue for RL applications. Existing works rely on extracting step-wise reward signals from trajectory-wise preference annotations, assuming that preferences correlate with the cumulative Markovian rewards. However, such methods fail to capture the holistic perspective of data annotation: Humans often assess the desirability of a sequence of actions by considering the overall outcome "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2407.04451","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/2407.04451/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":"2407.04451","created_at":"2026-07-05T08:40:33.847221+00:00"},{"alias_kind":"arxiv_version","alias_value":"2407.04451v1","created_at":"2026-07-05T08:40:33.847221+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2407.04451","created_at":"2026-07-05T08:40:33.847221+00:00"},{"alias_kind":"pith_short_12","alias_value":"TNJKNAHQXG37","created_at":"2026-07-05T08:40:33.847221+00:00"},{"alias_kind":"pith_short_16","alias_value":"TNJKNAHQXG37B7X5","created_at":"2026-07-05T08:40:33.847221+00:00"},{"alias_kind":"pith_short_8","alias_value":"TNJKNAHQ","created_at":"2026-07-05T08:40:33.847221+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.11982","citing_title":"PAWS: Preference Learning with Advantage-Weighted Segments","ref_index":48,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TNJKNAHQXG37B7X5S7STNVDBEQ","json":"https://pith.science/pith/TNJKNAHQXG37B7X5S7STNVDBEQ.json","graph_json":"https://pith.science/api/pith-number/TNJKNAHQXG37B7X5S7STNVDBEQ/graph.json","events_json":"https://pith.science/api/pith-number/TNJKNAHQXG37B7X5S7STNVDBEQ/events.json","paper":"https://pith.science/paper/TNJKNAHQ"},"agent_actions":{"view_html":"https://pith.science/pith/TNJKNAHQXG37B7X5S7STNVDBEQ","download_json":"https://pith.science/pith/TNJKNAHQXG37B7X5S7STNVDBEQ.json","view_paper":"https://pith.science/paper/TNJKNAHQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2407.04451&json=true","fetch_graph":"https://pith.science/api/pith-number/TNJKNAHQXG37B7X5S7STNVDBEQ/graph.json","fetch_events":"https://pith.science/api/pith-number/TNJKNAHQXG37B7X5S7STNVDBEQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TNJKNAHQXG37B7X5S7STNVDBEQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TNJKNAHQXG37B7X5S7STNVDBEQ/action/storage_attestation","attest_author":"https://pith.science/pith/TNJKNAHQXG37B7X5S7STNVDBEQ/action/author_attestation","sign_citation":"https://pith.science/pith/TNJKNAHQXG37B7X5S7STNVDBEQ/action/citation_signature","submit_replication":"https://pith.science/pith/TNJKNAHQXG37B7X5S7STNVDBEQ/action/replication_record"}},"created_at":"2026-07-05T08:40:33.847221+00:00","updated_at":"2026-07-05T08:40:33.847221+00:00"}