{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:ZBQKZPOR6BOWT7NMX3RVH4ES3B","short_pith_number":"pith:ZBQKZPOR","schema_version":"1.0","canonical_sha256":"c860acbdd1f05d69fdacbee353f092d867728f6f1937561f0bf5bb1365296520","source":{"kind":"arxiv","id":"1909.05236","version":2},"attestation_state":"computed","paper":{"title":"Safe Policy Improvement with an Estimated Baseline Policy","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"R\\'emi Tachet des Combes, Romain Laroche, Thiago D. Sim\\~ao","submitted_at":"2019-09-11T17:48:00Z","abstract_excerpt":"Previous work has shown the unreliability of existing algorithms in the batch Reinforcement Learning setting, and proposed the theoretically-grounded Safe Policy Improvement with Baseline Bootstrapping (SPIBB) fix: reproduce the baseline policy in the uncertain state-action pairs, in order to control the variance on the trained policy performance. However, in many real-world applications such as dialogue systems, pharmaceutical tests or crop management, data is collected under human supervision and the baseline remains unknown. In this paper, we apply SPIBB algorithms with a baseline estimate "},"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":"1909.05236","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2019-09-11T17:48:00Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"7a10b3b24b80a35eabccd29a939e90e7ff2d43cfe1219430b1a79d39f081d1b8","abstract_canon_sha256":"71c37d2af93238e6339d22d7cfca88eb5748f9f0d0644eec8ad287f81cddd8a8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:02:51.040914Z","signature_b64":"Zt8vZzid9h6wSR5c2n+DPoUU0Y00rREu1LBq2LJ1SX32tYaYhPwpSOJR8WaJ/E12bKptjomHmAvuaQuOC2kFAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c860acbdd1f05d69fdacbee353f092d867728f6f1937561f0bf5bb1365296520","last_reissued_at":"2026-07-05T02:02:51.040506Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:02:51.040506Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Safe Policy Improvement with an Estimated Baseline Policy","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"R\\'emi Tachet des Combes, Romain Laroche, Thiago D. Sim\\~ao","submitted_at":"2019-09-11T17:48:00Z","abstract_excerpt":"Previous work has shown the unreliability of existing algorithms in the batch Reinforcement Learning setting, and proposed the theoretically-grounded Safe Policy Improvement with Baseline Bootstrapping (SPIBB) fix: reproduce the baseline policy in the uncertain state-action pairs, in order to control the variance on the trained policy performance. However, in many real-world applications such as dialogue systems, pharmaceutical tests or crop management, data is collected under human supervision and the baseline remains unknown. In this paper, we apply SPIBB algorithms with a baseline estimate "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1909.05236","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/1909.05236/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":"1909.05236","created_at":"2026-07-05T02:02:51.040563+00:00"},{"alias_kind":"arxiv_version","alias_value":"1909.05236v2","created_at":"2026-07-05T02:02:51.040563+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1909.05236","created_at":"2026-07-05T02:02:51.040563+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZBQKZPOR6BOW","created_at":"2026-07-05T02:02:51.040563+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZBQKZPOR6BOWT7NM","created_at":"2026-07-05T02:02:51.040563+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZBQKZPOR","created_at":"2026-07-05T02:02:51.040563+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"1910.01708","citing_title":"Benchmarking Batch Deep Reinforcement Learning Algorithms","ref_index":13,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ZBQKZPOR6BOWT7NMX3RVH4ES3B","json":"https://pith.science/pith/ZBQKZPOR6BOWT7NMX3RVH4ES3B.json","graph_json":"https://pith.science/api/pith-number/ZBQKZPOR6BOWT7NMX3RVH4ES3B/graph.json","events_json":"https://pith.science/api/pith-number/ZBQKZPOR6BOWT7NMX3RVH4ES3B/events.json","paper":"https://pith.science/paper/ZBQKZPOR"},"agent_actions":{"view_html":"https://pith.science/pith/ZBQKZPOR6BOWT7NMX3RVH4ES3B","download_json":"https://pith.science/pith/ZBQKZPOR6BOWT7NMX3RVH4ES3B.json","view_paper":"https://pith.science/paper/ZBQKZPOR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1909.05236&json=true","fetch_graph":"https://pith.science/api/pith-number/ZBQKZPOR6BOWT7NMX3RVH4ES3B/graph.json","fetch_events":"https://pith.science/api/pith-number/ZBQKZPOR6BOWT7NMX3RVH4ES3B/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZBQKZPOR6BOWT7NMX3RVH4ES3B/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZBQKZPOR6BOWT7NMX3RVH4ES3B/action/storage_attestation","attest_author":"https://pith.science/pith/ZBQKZPOR6BOWT7NMX3RVH4ES3B/action/author_attestation","sign_citation":"https://pith.science/pith/ZBQKZPOR6BOWT7NMX3RVH4ES3B/action/citation_signature","submit_replication":"https://pith.science/pith/ZBQKZPOR6BOWT7NMX3RVH4ES3B/action/replication_record"}},"created_at":"2026-07-05T02:02:51.040563+00:00","updated_at":"2026-07-05T02:02:51.040563+00:00"}