{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:YF4JPHQGDU3O35P37HSNAGXM5B","short_pith_number":"pith:YF4JPHQG","schema_version":"1.0","canonical_sha256":"c178979e061d36edf5fbf9e4d01aece85b71c6cf064dbc34ed21dafe099e80f2","source":{"kind":"arxiv","id":"2605.18842","version":1},"attestation_state":"computed","paper":{"title":"Safe Continual Reinforcement Learning under Nonstationarity via Adaptive Safety Constraints","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Timofey Tomashevskiy","submitted_at":"2026-05-13T04:10:10Z","abstract_excerpt":"Safe reinforcement learning in nonstationary environments requires safety mechanisms that adapt as environmental conditions change. Standard safe reinforcement learning methods often assume fixed constraints or stable environmental conditions, which can become inadequate under distribution shift. We propose LILAC+, a framework for safe continual reinforcement learning under nonstationarity that combines three adaptive safety mechanisms: context-based safety constraints, adaptation-speed constraints, and budget-to-state safety enforcement. Context-based constraints adjust safety requirements us"},"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":"2605.18842","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2026-05-13T04:10:10Z","cross_cats_sorted":[],"title_canon_sha256":"30ab8a4474aa9a52574483579e625a197ed0253a32ec8f00ad207eb385d629be","abstract_canon_sha256":"94e2c908cf1cffc3ba3d4967fab64445c218d397e54291c5e5eaf7bd6dd4f70e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:06:25.288697Z","signature_b64":"2tOC2tBNaGlT2BEbKiv1k17Idqnv+77nMtzi0Vg1ob9dMXwXkw2LbonBd3CkRNUrQXDud7rxhPoWbM44sBxEBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c178979e061d36edf5fbf9e4d01aece85b71c6cf064dbc34ed21dafe099e80f2","last_reissued_at":"2026-05-20T00:06:25.287920Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:06:25.287920Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Safe Continual Reinforcement Learning under Nonstationarity via Adaptive Safety Constraints","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Timofey Tomashevskiy","submitted_at":"2026-05-13T04:10:10Z","abstract_excerpt":"Safe reinforcement learning in nonstationary environments requires safety mechanisms that adapt as environmental conditions change. Standard safe reinforcement learning methods often assume fixed constraints or stable environmental conditions, which can become inadequate under distribution shift. We propose LILAC+, a framework for safe continual reinforcement learning under nonstationarity that combines three adaptive safety mechanisms: context-based safety constraints, adaptation-speed constraints, and budget-to-state safety enforcement. Context-based constraints adjust safety requirements us"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.18842","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/2605.18842/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":"2605.18842","created_at":"2026-05-20T00:06:25.288045+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.18842v1","created_at":"2026-05-20T00:06:25.288045+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.18842","created_at":"2026-05-20T00:06:25.288045+00:00"},{"alias_kind":"pith_short_12","alias_value":"YF4JPHQGDU3O","created_at":"2026-05-20T00:06:25.288045+00:00"},{"alias_kind":"pith_short_16","alias_value":"YF4JPHQGDU3O35P3","created_at":"2026-05-20T00:06:25.288045+00:00"},{"alias_kind":"pith_short_8","alias_value":"YF4JPHQG","created_at":"2026-05-20T00:06:25.288045+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/YF4JPHQGDU3O35P37HSNAGXM5B","json":"https://pith.science/pith/YF4JPHQGDU3O35P37HSNAGXM5B.json","graph_json":"https://pith.science/api/pith-number/YF4JPHQGDU3O35P37HSNAGXM5B/graph.json","events_json":"https://pith.science/api/pith-number/YF4JPHQGDU3O35P37HSNAGXM5B/events.json","paper":"https://pith.science/paper/YF4JPHQG"},"agent_actions":{"view_html":"https://pith.science/pith/YF4JPHQGDU3O35P37HSNAGXM5B","download_json":"https://pith.science/pith/YF4JPHQGDU3O35P37HSNAGXM5B.json","view_paper":"https://pith.science/paper/YF4JPHQG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.18842&json=true","fetch_graph":"https://pith.science/api/pith-number/YF4JPHQGDU3O35P37HSNAGXM5B/graph.json","fetch_events":"https://pith.science/api/pith-number/YF4JPHQGDU3O35P37HSNAGXM5B/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YF4JPHQGDU3O35P37HSNAGXM5B/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YF4JPHQGDU3O35P37HSNAGXM5B/action/storage_attestation","attest_author":"https://pith.science/pith/YF4JPHQGDU3O35P37HSNAGXM5B/action/author_attestation","sign_citation":"https://pith.science/pith/YF4JPHQGDU3O35P37HSNAGXM5B/action/citation_signature","submit_replication":"https://pith.science/pith/YF4JPHQGDU3O35P37HSNAGXM5B/action/replication_record"}},"created_at":"2026-05-20T00:06:25.288045+00:00","updated_at":"2026-05-20T00:06:25.288045+00:00"}