{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2016:JDYAJTFSNC6AOIIIMVEM52FDSB","short_pith_number":"pith:JDYAJTFS","schema_version":"1.0","canonical_sha256":"48f004ccb268bc0721086548cee8a3906c9944a3f72b69164cd911e6cdd43b6e","source":{"kind":"arxiv","id":"1606.04753","version":2},"attestation_state":"computed","paper":{"title":"Safe Exploration in Finite Markov Decision Processes with Gaussian Processes","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.RO","stat.ML"],"primary_cat":"cs.LG","authors_text":"Andreas Krause, Felix Berkenkamp, Matteo Turchetta","submitted_at":"2016-06-15T13:18:30Z","abstract_excerpt":"In classical reinforcement learning, when exploring an environment, agents accept arbitrary short term loss for long term gain. This is infeasible for safety critical applications, such as robotics, where even a single unsafe action may cause system failure. In this paper, we address the problem of safely exploring finite Markov decision processes (MDP). We define safety in terms of an, a priori unknown, safety constraint that depends on states and actions. We aim to explore the MDP under this constraint, assuming that the unknown function satisfies regularity conditions expressed via a Gaussi"},"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":"1606.04753","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2016-06-15T13:18:30Z","cross_cats_sorted":["cs.AI","cs.RO","stat.ML"],"title_canon_sha256":"c24719c215c66532e5900e6ac6825ff194fb874fe432c4e641b6eb2de30df30a","abstract_canon_sha256":"087ec504a17ebc9b1996e3f5147b2298efd2ac04610e1b4002b6d09507462bd7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:52:02.022101Z","signature_b64":"3O6RE6cEdHLAk0giwA9UjepX5eDTSrXlHgro1RspYg+YZtfJ3QuPlmXOIQswSCRnlgMmLEKaWFeIQxcxpVDXDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"48f004ccb268bc0721086548cee8a3906c9944a3f72b69164cd911e6cdd43b6e","last_reissued_at":"2026-05-18T00:52:02.021566Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:52:02.021566Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Safe Exploration in Finite Markov Decision Processes with Gaussian Processes","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.RO","stat.ML"],"primary_cat":"cs.LG","authors_text":"Andreas Krause, Felix Berkenkamp, Matteo Turchetta","submitted_at":"2016-06-15T13:18:30Z","abstract_excerpt":"In classical reinforcement learning, when exploring an environment, agents accept arbitrary short term loss for long term gain. This is infeasible for safety critical applications, such as robotics, where even a single unsafe action may cause system failure. In this paper, we address the problem of safely exploring finite Markov decision processes (MDP). We define safety in terms of an, a priori unknown, safety constraint that depends on states and actions. We aim to explore the MDP under this constraint, assuming that the unknown function satisfies regularity conditions expressed via a Gaussi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1606.04753","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":""},"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":"1606.04753","created_at":"2026-05-18T00:52:02.021650+00:00"},{"alias_kind":"arxiv_version","alias_value":"1606.04753v2","created_at":"2026-05-18T00:52:02.021650+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1606.04753","created_at":"2026-05-18T00:52:02.021650+00:00"},{"alias_kind":"pith_short_12","alias_value":"JDYAJTFSNC6A","created_at":"2026-05-18T12:30:25.849896+00:00"},{"alias_kind":"pith_short_16","alias_value":"JDYAJTFSNC6AOIII","created_at":"2026-05-18T12:30:25.849896+00:00"},{"alias_kind":"pith_short_8","alias_value":"JDYAJTFS","created_at":"2026-05-18T12:30:25.849896+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"1606.06565","citing_title":"Concrete Problems in AI Safety","ref_index":160,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JDYAJTFSNC6AOIIIMVEM52FDSB","json":"https://pith.science/pith/JDYAJTFSNC6AOIIIMVEM52FDSB.json","graph_json":"https://pith.science/api/pith-number/JDYAJTFSNC6AOIIIMVEM52FDSB/graph.json","events_json":"https://pith.science/api/pith-number/JDYAJTFSNC6AOIIIMVEM52FDSB/events.json","paper":"https://pith.science/paper/JDYAJTFS"},"agent_actions":{"view_html":"https://pith.science/pith/JDYAJTFSNC6AOIIIMVEM52FDSB","download_json":"https://pith.science/pith/JDYAJTFSNC6AOIIIMVEM52FDSB.json","view_paper":"https://pith.science/paper/JDYAJTFS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1606.04753&json=true","fetch_graph":"https://pith.science/api/pith-number/JDYAJTFSNC6AOIIIMVEM52FDSB/graph.json","fetch_events":"https://pith.science/api/pith-number/JDYAJTFSNC6AOIIIMVEM52FDSB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JDYAJTFSNC6AOIIIMVEM52FDSB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JDYAJTFSNC6AOIIIMVEM52FDSB/action/storage_attestation","attest_author":"https://pith.science/pith/JDYAJTFSNC6AOIIIMVEM52FDSB/action/author_attestation","sign_citation":"https://pith.science/pith/JDYAJTFSNC6AOIIIMVEM52FDSB/action/citation_signature","submit_replication":"https://pith.science/pith/JDYAJTFSNC6AOIIIMVEM52FDSB/action/replication_record"}},"created_at":"2026-05-18T00:52:02.021650+00:00","updated_at":"2026-05-18T00:52:02.021650+00:00"}