{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:62SZW72YDARDYLWB6LIYSCKEGM","short_pith_number":"pith:62SZW72Y","schema_version":"1.0","canonical_sha256":"f6a59b7f5818223c2ec1f2d1890944330180042c9cf112abd5cdead44b229afe","source":{"kind":"arxiv","id":"1904.06387","version":5},"attestation_state":"computed","paper":{"title":"Extrapolating Beyond Suboptimal Demonstrations via Inverse Reinforcement Learning from Observations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Daniel S. Brown, Prabhat Nagarajan, Scott Niekum, Wonjoon Goo","submitted_at":"2019-04-12T19:34:43Z","abstract_excerpt":"A critical flaw of existing inverse reinforcement learning (IRL) methods is their inability to significantly outperform the demonstrator. This is because IRL typically seeks a reward function that makes the demonstrator appear near-optimal, rather than inferring the underlying intentions of the demonstrator that may have been poorly executed in practice. In this paper, we introduce a novel reward-learning-from-observation algorithm, Trajectory-ranked Reward EXtrapolation (T-REX), that extrapolates beyond a set of (approximately) ranked demonstrations in order to infer high-quality reward funct"},"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":"1904.06387","kind":"arxiv","version":5},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-04-12T19:34:43Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"94629d2e5117df8269e8db35152cd2f6070d55cad20c750b041949446ff8c999","abstract_canon_sha256":"21196885823e6a131448a971742824da7ccfb033bcbefbf107582c3b087e6f1c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:41:09.056486Z","signature_b64":"oJGjZ6eUsgpEIwxzuh3KLzBkQSF7IUfDwmKdhhcZum9ReTQmUmIoIpCkMXQrQVxJvuT2GqJEOHkWgPvY25cAAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f6a59b7f5818223c2ec1f2d1890944330180042c9cf112abd5cdead44b229afe","last_reissued_at":"2026-05-17T23:41:09.055783Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:41:09.055783Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Extrapolating Beyond Suboptimal Demonstrations via Inverse Reinforcement Learning from Observations","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Daniel S. Brown, Prabhat Nagarajan, Scott Niekum, Wonjoon Goo","submitted_at":"2019-04-12T19:34:43Z","abstract_excerpt":"A critical flaw of existing inverse reinforcement learning (IRL) methods is their inability to significantly outperform the demonstrator. This is because IRL typically seeks a reward function that makes the demonstrator appear near-optimal, rather than inferring the underlying intentions of the demonstrator that may have been poorly executed in practice. In this paper, we introduce a novel reward-learning-from-observation algorithm, Trajectory-ranked Reward EXtrapolation (T-REX), that extrapolates beyond a set of (approximately) ranked demonstrations in order to infer high-quality reward funct"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1904.06387","kind":"arxiv","version":5},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"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":"1904.06387","created_at":"2026-05-17T23:41:09.055904+00:00"},{"alias_kind":"arxiv_version","alias_value":"1904.06387v5","created_at":"2026-05-17T23:41:09.055904+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1904.06387","created_at":"2026-05-17T23:41:09.055904+00:00"},{"alias_kind":"pith_short_12","alias_value":"62SZW72YDARD","created_at":"2026-05-18T12:33:10.108867+00:00"},{"alias_kind":"pith_short_16","alias_value":"62SZW72YDARDYLWB","created_at":"2026-05-18T12:33:10.108867+00:00"},{"alias_kind":"pith_short_8","alias_value":"62SZW72Y","created_at":"2026-05-18T12:33:10.108867+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2409.12917","citing_title":"Training Language Models to Self-Correct via Reinforcement Learning","ref_index":219,"is_internal_anchor":true},{"citing_arxiv_id":"2112.00861","citing_title":"A General Language Assistant as a Laboratory for Alignment","ref_index":211,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/62SZW72YDARDYLWB6LIYSCKEGM","json":"https://pith.science/pith/62SZW72YDARDYLWB6LIYSCKEGM.json","graph_json":"https://pith.science/api/pith-number/62SZW72YDARDYLWB6LIYSCKEGM/graph.json","events_json":"https://pith.science/api/pith-number/62SZW72YDARDYLWB6LIYSCKEGM/events.json","paper":"https://pith.science/paper/62SZW72Y"},"agent_actions":{"view_html":"https://pith.science/pith/62SZW72YDARDYLWB6LIYSCKEGM","download_json":"https://pith.science/pith/62SZW72YDARDYLWB6LIYSCKEGM.json","view_paper":"https://pith.science/paper/62SZW72Y","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1904.06387&json=true","fetch_graph":"https://pith.science/api/pith-number/62SZW72YDARDYLWB6LIYSCKEGM/graph.json","fetch_events":"https://pith.science/api/pith-number/62SZW72YDARDYLWB6LIYSCKEGM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/62SZW72YDARDYLWB6LIYSCKEGM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/62SZW72YDARDYLWB6LIYSCKEGM/action/storage_attestation","attest_author":"https://pith.science/pith/62SZW72YDARDYLWB6LIYSCKEGM/action/author_attestation","sign_citation":"https://pith.science/pith/62SZW72YDARDYLWB6LIYSCKEGM/action/citation_signature","submit_replication":"https://pith.science/pith/62SZW72YDARDYLWB6LIYSCKEGM/action/replication_record"}},"created_at":"2026-05-17T23:41:09.055904+00:00","updated_at":"2026-05-17T23:41:09.055904+00:00"}