{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:WEJ3KGHCSAWCOMF7DO4OCCBXYV","short_pith_number":"pith:WEJ3KGHC","schema_version":"1.0","canonical_sha256":"b113b518e2902c2730bf1bb8e10837c57d9d7effda176739258dff27b2f16508","source":{"kind":"arxiv","id":"1807.06096","version":2},"attestation_state":"computed","paper":{"title":"Safe Reinforcement Learning via Probabilistic Shields","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Alexandru C. Serban, Bettina K\\\"onighofer, Nils Jansen, Roderick Bloem, Sebastian Junges","submitted_at":"2018-07-16T20:29:04Z","abstract_excerpt":"This paper targets the efficient construction of a safety shield for decision making in scenarios that incorporate uncertainty. Markov decision processes (MDPs) are prominent models to capture such planning problems. Reinforcement learning (RL) is a machine learning technique to determine near-optimal policies in MDPs that may be unknown prior to exploring the model. However, during exploration, RL is prone to induce behavior that is undesirable or not allowed in safety- or mission-critical contexts. We introduce the concept of a probabilistic shield that enables decision-making to adhere to s"},"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":"1807.06096","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2018-07-16T20:29:04Z","cross_cats_sorted":[],"title_canon_sha256":"e1e8db47da002e7db3ef0d1c2413692e99f545580659040514d3210abc5f92ca","abstract_canon_sha256":"3e19b4a81f6b82f3abf7cb7ce0256bb428b05705ffdd88afc9d8b4a63d094930"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:21:31.365081Z","signature_b64":"HNkjayhaVHb0XYTyHL8eJOiKd9mYfSxrNZjVA8zXraebwq4/i05hAJaLayw6IK3Lb8M6MnhvKnw7O12PgGxqAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b113b518e2902c2730bf1bb8e10837c57d9d7effda176739258dff27b2f16508","last_reissued_at":"2026-07-05T00:21:31.364624Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:21:31.364624Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Safe Reinforcement Learning via Probabilistic Shields","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Alexandru C. Serban, Bettina K\\\"onighofer, Nils Jansen, Roderick Bloem, Sebastian Junges","submitted_at":"2018-07-16T20:29:04Z","abstract_excerpt":"This paper targets the efficient construction of a safety shield for decision making in scenarios that incorporate uncertainty. Markov decision processes (MDPs) are prominent models to capture such planning problems. Reinforcement learning (RL) is a machine learning technique to determine near-optimal policies in MDPs that may be unknown prior to exploring the model. However, during exploration, RL is prone to induce behavior that is undesirable or not allowed in safety- or mission-critical contexts. We introduce the concept of a probabilistic shield that enables decision-making to adhere to s"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1807.06096","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/1807.06096/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":"1807.06096","created_at":"2026-07-05T00:21:31.364681+00:00"},{"alias_kind":"arxiv_version","alias_value":"1807.06096v2","created_at":"2026-07-05T00:21:31.364681+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1807.06096","created_at":"2026-07-05T00:21:31.364681+00:00"},{"alias_kind":"pith_short_12","alias_value":"WEJ3KGHCSAWC","created_at":"2026-07-05T00:21:31.364681+00:00"},{"alias_kind":"pith_short_16","alias_value":"WEJ3KGHCSAWCOMF7","created_at":"2026-07-05T00:21:31.364681+00:00"},{"alias_kind":"pith_short_8","alias_value":"WEJ3KGHC","created_at":"2026-07-05T00:21:31.364681+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.08312","citing_title":"Neuro-Symbolic Injection of LTLf Constraints in Autoregressive Reinforcement Learning Policies","ref_index":35,"is_internal_anchor":false},{"citing_arxiv_id":"2606.31023","citing_title":"Certified Speculative Execution for Untrusted AI Agents","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2603.16651","citing_title":"What if Pinocchio Were a Reinforcement Learning Agent: A Normative End-to-End Pipeline","ref_index":98,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/WEJ3KGHCSAWCOMF7DO4OCCBXYV","json":"https://pith.science/pith/WEJ3KGHCSAWCOMF7DO4OCCBXYV.json","graph_json":"https://pith.science/api/pith-number/WEJ3KGHCSAWCOMF7DO4OCCBXYV/graph.json","events_json":"https://pith.science/api/pith-number/WEJ3KGHCSAWCOMF7DO4OCCBXYV/events.json","paper":"https://pith.science/paper/WEJ3KGHC"},"agent_actions":{"view_html":"https://pith.science/pith/WEJ3KGHCSAWCOMF7DO4OCCBXYV","download_json":"https://pith.science/pith/WEJ3KGHCSAWCOMF7DO4OCCBXYV.json","view_paper":"https://pith.science/paper/WEJ3KGHC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1807.06096&json=true","fetch_graph":"https://pith.science/api/pith-number/WEJ3KGHCSAWCOMF7DO4OCCBXYV/graph.json","fetch_events":"https://pith.science/api/pith-number/WEJ3KGHCSAWCOMF7DO4OCCBXYV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/WEJ3KGHCSAWCOMF7DO4OCCBXYV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/WEJ3KGHCSAWCOMF7DO4OCCBXYV/action/storage_attestation","attest_author":"https://pith.science/pith/WEJ3KGHCSAWCOMF7DO4OCCBXYV/action/author_attestation","sign_citation":"https://pith.science/pith/WEJ3KGHCSAWCOMF7DO4OCCBXYV/action/citation_signature","submit_replication":"https://pith.science/pith/WEJ3KGHCSAWCOMF7DO4OCCBXYV/action/replication_record"}},"created_at":"2026-07-05T00:21:31.364681+00:00","updated_at":"2026-07-05T00:21:31.364681+00:00"}