{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:P3CUP4WATGNCZBYMU25ZKNY56D","short_pith_number":"pith:P3CUP4WA","schema_version":"1.0","canonical_sha256":"7ec547f2c0999a2c870ca6bb95371df0d315af6a07680e370a8581aa63fa0005","source":{"kind":"arxiv","id":"2506.06959","version":1},"attestation_state":"computed","paper":{"title":"Deontically Constrained Policy Improvement in Reinforcement Learning Agents","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Alena Makarova, Houssam Abbas","submitted_at":"2025-06-08T01:01:06Z","abstract_excerpt":"Markov Decision Processes (MDPs) are the most common model for decision making under uncertainty in the Machine Learning community. An MDP captures non-determinism, probabilistic uncertainty, and an explicit model of action. A Reinforcement Learning (RL) agent learns to act in an MDP by maximizing a utility function. This paper considers the problem of learning a decision policy that maximizes utility subject to satisfying a constraint expressed in deontic logic. In this setup, the utility captures the agent's mission - such as going quickly from A to B. The deontic formula represents (ethical"},"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":"2506.06959","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.AI","submitted_at":"2025-06-08T01:01:06Z","cross_cats_sorted":[],"title_canon_sha256":"00f8d7b24e3e67a80cad527ce1d190b6f938ea657d576c3691aef1f003b6f259","abstract_canon_sha256":"db89610742283ee16df9a377d8f8a628ff99961a0aaf6127bef76ea967f929d9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:17:58.776844Z","signature_b64":"jfarUd0qd4IpnmSSy/xs7tZ3ligSx2zj8xE6eSzPUuy2Kw7f8jz3QIYmocm+zqDyxBPG6ifiYs8K611G9tU9AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7ec547f2c0999a2c870ca6bb95371df0d315af6a07680e370a8581aa63fa0005","last_reissued_at":"2026-07-05T11:17:58.776353Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:17:58.776353Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Deontically Constrained Policy Improvement in Reinforcement Learning Agents","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Alena Makarova, Houssam Abbas","submitted_at":"2025-06-08T01:01:06Z","abstract_excerpt":"Markov Decision Processes (MDPs) are the most common model for decision making under uncertainty in the Machine Learning community. An MDP captures non-determinism, probabilistic uncertainty, and an explicit model of action. A Reinforcement Learning (RL) agent learns to act in an MDP by maximizing a utility function. This paper considers the problem of learning a decision policy that maximizes utility subject to satisfying a constraint expressed in deontic logic. In this setup, the utility captures the agent's mission - such as going quickly from A to B. The deontic formula represents (ethical"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.06959","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/2506.06959/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":"2506.06959","created_at":"2026-07-05T11:17:58.776418+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.06959v1","created_at":"2026-07-05T11:17:58.776418+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.06959","created_at":"2026-07-05T11:17:58.776418+00:00"},{"alias_kind":"pith_short_12","alias_value":"P3CUP4WATGNC","created_at":"2026-07-05T11:17:58.776418+00:00"},{"alias_kind":"pith_short_16","alias_value":"P3CUP4WATGNCZBYM","created_at":"2026-07-05T11:17:58.776418+00:00"},{"alias_kind":"pith_short_8","alias_value":"P3CUP4WA","created_at":"2026-07-05T11:17:58.776418+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/P3CUP4WATGNCZBYMU25ZKNY56D","json":"https://pith.science/pith/P3CUP4WATGNCZBYMU25ZKNY56D.json","graph_json":"https://pith.science/api/pith-number/P3CUP4WATGNCZBYMU25ZKNY56D/graph.json","events_json":"https://pith.science/api/pith-number/P3CUP4WATGNCZBYMU25ZKNY56D/events.json","paper":"https://pith.science/paper/P3CUP4WA"},"agent_actions":{"view_html":"https://pith.science/pith/P3CUP4WATGNCZBYMU25ZKNY56D","download_json":"https://pith.science/pith/P3CUP4WATGNCZBYMU25ZKNY56D.json","view_paper":"https://pith.science/paper/P3CUP4WA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.06959&json=true","fetch_graph":"https://pith.science/api/pith-number/P3CUP4WATGNCZBYMU25ZKNY56D/graph.json","fetch_events":"https://pith.science/api/pith-number/P3CUP4WATGNCZBYMU25ZKNY56D/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/P3CUP4WATGNCZBYMU25ZKNY56D/action/timestamp_anchor","attest_storage":"https://pith.science/pith/P3CUP4WATGNCZBYMU25ZKNY56D/action/storage_attestation","attest_author":"https://pith.science/pith/P3CUP4WATGNCZBYMU25ZKNY56D/action/author_attestation","sign_citation":"https://pith.science/pith/P3CUP4WATGNCZBYMU25ZKNY56D/action/citation_signature","submit_replication":"https://pith.science/pith/P3CUP4WATGNCZBYMU25ZKNY56D/action/replication_record"}},"created_at":"2026-07-05T11:17:58.776418+00:00","updated_at":"2026-07-05T11:17:58.776418+00:00"}