{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:IGVJ4GBZVOFH5VHT7VNZDC3QCR","short_pith_number":"pith:IGVJ4GBZ","schema_version":"1.0","canonical_sha256":"41aa9e1839ab8a7ed4f3fd5b918b7014546b1379e30dbc1a431a6f667f09989f","source":{"kind":"arxiv","id":"2605.18841","version":1},"attestation_state":"computed","paper":{"title":"From Cumulative Constraints to Adaptive Runtime Safety Control for Nonstationary Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Timofey Tomashevskiy","submitted_at":"2026-05-13T03:34:13Z","abstract_excerpt":"Safety in reinforcement learning is often specified through cumulative cost constraints, but these trajectory-level guarantees do not directly prevent unsafe individual decisions, especially under nonstationarity. In continual and nonstationary settings, the difficulty is amplified because the risk associated with the same action can vary across contexts, while a fixed state-level threshold may be either too conservative or too weak. We propose Constraint Projection Safety Shield (CPSS), a runtime mechanism that converts a cumulative safety budget into adaptive state-level control constraints "},"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":"2605.18841","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T03:34:13Z","cross_cats_sorted":[],"title_canon_sha256":"67c20f15e882c4960695354512a35cc4722c89962ac1adb888e359e5e6af6c5a","abstract_canon_sha256":"7aa4da326cb389f1846a876b4d869155c50b8eb0f6972604df14e9df344237a9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:06:25.265560Z","signature_b64":"FIBAqrLzTYk/lfYlVGdisWpkxGGCl56Hz+FBn+aqX307DuFLAKVRRL3jPnWKv48H6ihWS75/SWBSy4Rr0PRtBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"41aa9e1839ab8a7ed4f3fd5b918b7014546b1379e30dbc1a431a6f667f09989f","last_reissued_at":"2026-05-20T00:06:25.264772Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:06:25.264772Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"From Cumulative Constraints to Adaptive Runtime Safety Control for Nonstationary Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Timofey Tomashevskiy","submitted_at":"2026-05-13T03:34:13Z","abstract_excerpt":"Safety in reinforcement learning is often specified through cumulative cost constraints, but these trajectory-level guarantees do not directly prevent unsafe individual decisions, especially under nonstationarity. In continual and nonstationary settings, the difficulty is amplified because the risk associated with the same action can vary across contexts, while a fixed state-level threshold may be either too conservative or too weak. We propose Constraint Projection Safety Shield (CPSS), a runtime mechanism that converts a cumulative safety budget into adaptive state-level control constraints "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.18841","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/2605.18841/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":"2605.18841","created_at":"2026-05-20T00:06:25.264906+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.18841v1","created_at":"2026-05-20T00:06:25.264906+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.18841","created_at":"2026-05-20T00:06:25.264906+00:00"},{"alias_kind":"pith_short_12","alias_value":"IGVJ4GBZVOFH","created_at":"2026-05-20T00:06:25.264906+00:00"},{"alias_kind":"pith_short_16","alias_value":"IGVJ4GBZVOFH5VHT","created_at":"2026-05-20T00:06:25.264906+00:00"},{"alias_kind":"pith_short_8","alias_value":"IGVJ4GBZ","created_at":"2026-05-20T00:06:25.264906+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/IGVJ4GBZVOFH5VHT7VNZDC3QCR","json":"https://pith.science/pith/IGVJ4GBZVOFH5VHT7VNZDC3QCR.json","graph_json":"https://pith.science/api/pith-number/IGVJ4GBZVOFH5VHT7VNZDC3QCR/graph.json","events_json":"https://pith.science/api/pith-number/IGVJ4GBZVOFH5VHT7VNZDC3QCR/events.json","paper":"https://pith.science/paper/IGVJ4GBZ"},"agent_actions":{"view_html":"https://pith.science/pith/IGVJ4GBZVOFH5VHT7VNZDC3QCR","download_json":"https://pith.science/pith/IGVJ4GBZVOFH5VHT7VNZDC3QCR.json","view_paper":"https://pith.science/paper/IGVJ4GBZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.18841&json=true","fetch_graph":"https://pith.science/api/pith-number/IGVJ4GBZVOFH5VHT7VNZDC3QCR/graph.json","fetch_events":"https://pith.science/api/pith-number/IGVJ4GBZVOFH5VHT7VNZDC3QCR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IGVJ4GBZVOFH5VHT7VNZDC3QCR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IGVJ4GBZVOFH5VHT7VNZDC3QCR/action/storage_attestation","attest_author":"https://pith.science/pith/IGVJ4GBZVOFH5VHT7VNZDC3QCR/action/author_attestation","sign_citation":"https://pith.science/pith/IGVJ4GBZVOFH5VHT7VNZDC3QCR/action/citation_signature","submit_replication":"https://pith.science/pith/IGVJ4GBZVOFH5VHT7VNZDC3QCR/action/replication_record"}},"created_at":"2026-05-20T00:06:25.264906+00:00","updated_at":"2026-05-20T00:06:25.264906+00:00"}