{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:MPAZKAFH3YHK22EJPUN6H3BKIL","short_pith_number":"pith:MPAZKAFH","schema_version":"1.0","canonical_sha256":"63c19500a7de0ead68897d1be3ec2a42cea64b3be766c0771cff9bb25d75067f","source":{"kind":"arxiv","id":"2405.15994","version":2},"attestation_state":"computed","paper":{"title":"Verified Safe Reinforcement Learning for Neural Network Dynamic Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Huan Zhang, Junlin Wu, Yevgeniy Vorobeychik","submitted_at":"2024-05-25T00:35:39Z","abstract_excerpt":"Learning reliably safe autonomous control is one of the core problems in trustworthy autonomy. However, training a controller that can be formally verified to be safe remains a major challenge. We introduce a novel approach for learning verified safe control policies in nonlinear neural dynamical systems while maximizing overall performance. Our approach aims to achieve safety in the sense of finite-horizon reachability proofs, and is comprised of three key parts. The first is a novel curriculum learning scheme that iteratively increases the verified safe horizon. The second leverages the iter"},"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":"2405.15994","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-05-25T00:35:39Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"371c5216790f090466429d8c716f461f3a0f3558d2d4af01fd2c0d1906d05463","abstract_canon_sha256":"02a2e0afdb3193f7b4351ba43f4b7d872ded4d7c0cea718cee74726bbe8c041b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:36:11.717743Z","signature_b64":"MnfLqn3nJR6aOGq16ajRqDcPYNLQ/NcwZ6NDZjAfG6R7LJmkZLf+z5KmuT8JkswV3ptszRaFY1ySETHF3+N7Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"63c19500a7de0ead68897d1be3ec2a42cea64b3be766c0771cff9bb25d75067f","last_reissued_at":"2026-07-05T09:36:11.717070Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:36:11.717070Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Verified Safe Reinforcement Learning for Neural Network Dynamic Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Huan Zhang, Junlin Wu, Yevgeniy Vorobeychik","submitted_at":"2024-05-25T00:35:39Z","abstract_excerpt":"Learning reliably safe autonomous control is one of the core problems in trustworthy autonomy. However, training a controller that can be formally verified to be safe remains a major challenge. We introduce a novel approach for learning verified safe control policies in nonlinear neural dynamical systems while maximizing overall performance. Our approach aims to achieve safety in the sense of finite-horizon reachability proofs, and is comprised of three key parts. The first is a novel curriculum learning scheme that iteratively increases the verified safe horizon. The second leverages the iter"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.15994","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/2405.15994/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":"2405.15994","created_at":"2026-07-05T09:36:11.717160+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.15994v2","created_at":"2026-07-05T09:36:11.717160+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.15994","created_at":"2026-07-05T09:36:11.717160+00:00"},{"alias_kind":"pith_short_12","alias_value":"MPAZKAFH3YHK","created_at":"2026-07-05T09:36:11.717160+00:00"},{"alias_kind":"pith_short_16","alias_value":"MPAZKAFH3YHK22EJ","created_at":"2026-07-05T09:36:11.717160+00:00"},{"alias_kind":"pith_short_8","alias_value":"MPAZKAFH","created_at":"2026-07-05T09:36:11.717160+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2412.09584","citing_title":"BaB-ND: Long-Horizon Motion Planning with Branch-and-Bound and Neural Dynamics","ref_index":63,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MPAZKAFH3YHK22EJPUN6H3BKIL","json":"https://pith.science/pith/MPAZKAFH3YHK22EJPUN6H3BKIL.json","graph_json":"https://pith.science/api/pith-number/MPAZKAFH3YHK22EJPUN6H3BKIL/graph.json","events_json":"https://pith.science/api/pith-number/MPAZKAFH3YHK22EJPUN6H3BKIL/events.json","paper":"https://pith.science/paper/MPAZKAFH"},"agent_actions":{"view_html":"https://pith.science/pith/MPAZKAFH3YHK22EJPUN6H3BKIL","download_json":"https://pith.science/pith/MPAZKAFH3YHK22EJPUN6H3BKIL.json","view_paper":"https://pith.science/paper/MPAZKAFH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.15994&json=true","fetch_graph":"https://pith.science/api/pith-number/MPAZKAFH3YHK22EJPUN6H3BKIL/graph.json","fetch_events":"https://pith.science/api/pith-number/MPAZKAFH3YHK22EJPUN6H3BKIL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MPAZKAFH3YHK22EJPUN6H3BKIL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MPAZKAFH3YHK22EJPUN6H3BKIL/action/storage_attestation","attest_author":"https://pith.science/pith/MPAZKAFH3YHK22EJPUN6H3BKIL/action/author_attestation","sign_citation":"https://pith.science/pith/MPAZKAFH3YHK22EJPUN6H3BKIL/action/citation_signature","submit_replication":"https://pith.science/pith/MPAZKAFH3YHK22EJPUN6H3BKIL/action/replication_record"}},"created_at":"2026-07-05T09:36:11.717160+00:00","updated_at":"2026-07-05T09:36:11.717160+00:00"}