{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2022:IROQ4KK2GPAMTHMD4CXKYHOJ3V","short_pith_number":"pith:IROQ4KK2","canonical_record":{"source":{"id":"2209.07089","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-09-15T07:01:42Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"ae6ccd5d7ec90533e26ccb8ec4057b503eef57d825aff261c9c77a85283de73b","abstract_canon_sha256":"757a7b1feaa45b0cefa0d7b1810a32aff79960827ee826ee2af43209f57ae01f"},"schema_version":"1.0"},"canonical_sha256":"445d0e295a33c0c99d83e0aeac1dc9dd7bb45e9148d2aee00779fcf32d7313fe","source":{"kind":"arxiv","id":"2209.07089","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2209.07089","created_at":"2026-07-05T05:14:36Z"},{"alias_kind":"arxiv_version","alias_value":"2209.07089v2","created_at":"2026-07-05T05:14:36Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2209.07089","created_at":"2026-07-05T05:14:36Z"},{"alias_kind":"pith_short_12","alias_value":"IROQ4KK2GPAM","created_at":"2026-07-05T05:14:36Z"},{"alias_kind":"pith_short_16","alias_value":"IROQ4KK2GPAMTHMD","created_at":"2026-07-05T05:14:36Z"},{"alias_kind":"pith_short_8","alias_value":"IROQ4KK2","created_at":"2026-07-05T05:14:36Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2022:IROQ4KK2GPAMTHMD4CXKYHOJ3V","target":"record","payload":{"canonical_record":{"source":{"id":"2209.07089","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-09-15T07:01:42Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"ae6ccd5d7ec90533e26ccb8ec4057b503eef57d825aff261c9c77a85283de73b","abstract_canon_sha256":"757a7b1feaa45b0cefa0d7b1810a32aff79960827ee826ee2af43209f57ae01f"},"schema_version":"1.0"},"canonical_sha256":"445d0e295a33c0c99d83e0aeac1dc9dd7bb45e9148d2aee00779fcf32d7313fe","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:14:36.598262Z","signature_b64":"oBhrdaO88yVKNvEL7LCApE7KzGVj+6nSsN1or+A1hOcUrrHyjt80776Owaspeq/330XXTozwggKRhKhQehKIAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"445d0e295a33c0c99d83e0aeac1dc9dd7bb45e9148d2aee00779fcf32d7313fe","last_reissued_at":"2026-07-05T05:14:36.597725Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:14:36.597725Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2209.07089","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T05:14:36Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"U68zFQQPlqGKYWyL4fRB/4v3y8LLeRR5vR7zuz2OW9MLsLYOHkl7UGr0Zp/kmuChhcdyh10AvuFsLFMIw4bfDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-09T06:01:31.120282Z"},"content_sha256":"90cd10a8e84e879f0891fed435898055382495d48fd6345c47f172e4419c974b","schema_version":"1.0","event_id":"sha256:90cd10a8e84e879f0891fed435898055382495d48fd6345c47f172e4419c974b"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2022:IROQ4KK2GPAMTHMD4CXKYHOJ3V","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Constrained Update Projection Approach to Safe Policy Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Binbin Zhou, Gang Pan, Jiaming Ji, Juntao Dai, Linrui Zhang, Long Yang, Pengfei Li, Yaodong Yang","submitted_at":"2022-09-15T07:01:42Z","abstract_excerpt":"Safe reinforcement learning (RL) studies problems where an intelligent agent has to not only maximize reward but also avoid exploring unsafe areas. In this study, we propose CUP, a novel policy optimization method based on Constrained Update Projection framework that enjoys rigorous safety guarantee. Central to our CUP development is the newly proposed surrogate functions along with the performance bound. Compared to previous safe RL methods, CUP enjoys the benefits of 1) CUP generalizes the surrogate functions to generalized advantage estimator (GAE), leading to strong empirical performance. "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2209.07089","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/2209.07089/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T05:14:36Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Pj5OJ+oC/3iW3lA9FLoGQ6aVk8vboun7TyhfSo0ZRigdrd+4Gq6hhqguDVXx9Xhbj4tHlSdcUdOYLo/Z94GGBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-09T06:01:31.120737Z"},"content_sha256":"aabef6f178258e5d12306a092c427bfe19012949c9a548181a05375c86c998f5","schema_version":"1.0","event_id":"sha256:aabef6f178258e5d12306a092c427bfe19012949c9a548181a05375c86c998f5"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/IROQ4KK2GPAMTHMD4CXKYHOJ3V/bundle.json","state_url":"https://pith.science/pith/IROQ4KK2GPAMTHMD4CXKYHOJ3V/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/IROQ4KK2GPAMTHMD4CXKYHOJ3V/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-07-09T06:01:31Z","links":{"resolver":"https://pith.science/pith/IROQ4KK2GPAMTHMD4CXKYHOJ3V","bundle":"https://pith.science/pith/IROQ4KK2GPAMTHMD4CXKYHOJ3V/bundle.json","state":"https://pith.science/pith/IROQ4KK2GPAMTHMD4CXKYHOJ3V/state.json","well_known_bundle":"https://pith.science/.well-known/pith/IROQ4KK2GPAMTHMD4CXKYHOJ3V/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:IROQ4KK2GPAMTHMD4CXKYHOJ3V","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"757a7b1feaa45b0cefa0d7b1810a32aff79960827ee826ee2af43209f57ae01f","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-09-15T07:01:42Z","title_canon_sha256":"ae6ccd5d7ec90533e26ccb8ec4057b503eef57d825aff261c9c77a85283de73b"},"schema_version":"1.0","source":{"id":"2209.07089","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2209.07089","created_at":"2026-07-05T05:14:36Z"},{"alias_kind":"arxiv_version","alias_value":"2209.07089v2","created_at":"2026-07-05T05:14:36Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2209.07089","created_at":"2026-07-05T05:14:36Z"},{"alias_kind":"pith_short_12","alias_value":"IROQ4KK2GPAM","created_at":"2026-07-05T05:14:36Z"},{"alias_kind":"pith_short_16","alias_value":"IROQ4KK2GPAMTHMD","created_at":"2026-07-05T05:14:36Z"},{"alias_kind":"pith_short_8","alias_value":"IROQ4KK2","created_at":"2026-07-05T05:14:36Z"}],"graph_snapshots":[{"event_id":"sha256:aabef6f178258e5d12306a092c427bfe19012949c9a548181a05375c86c998f5","target":"graph","created_at":"2026-07-05T05:14:36Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2209.07089/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Safe reinforcement learning (RL) studies problems where an intelligent agent has to not only maximize reward but also avoid exploring unsafe areas. In this study, we propose CUP, a novel policy optimization method based on Constrained Update Projection framework that enjoys rigorous safety guarantee. Central to our CUP development is the newly proposed surrogate functions along with the performance bound. Compared to previous safe RL methods, CUP enjoys the benefits of 1) CUP generalizes the surrogate functions to generalized advantage estimator (GAE), leading to strong empirical performance. ","authors_text":"Binbin Zhou, Gang Pan, Jiaming Ji, Juntao Dai, Linrui Zhang, Long Yang, Pengfei Li, Yaodong Yang","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-09-15T07:01:42Z","title":"Constrained Update Projection Approach to Safe Policy Optimization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2209.07089","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:90cd10a8e84e879f0891fed435898055382495d48fd6345c47f172e4419c974b","target":"record","created_at":"2026-07-05T05:14:36Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"757a7b1feaa45b0cefa0d7b1810a32aff79960827ee826ee2af43209f57ae01f","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-09-15T07:01:42Z","title_canon_sha256":"ae6ccd5d7ec90533e26ccb8ec4057b503eef57d825aff261c9c77a85283de73b"},"schema_version":"1.0","source":{"id":"2209.07089","kind":"arxiv","version":2}},"canonical_sha256":"445d0e295a33c0c99d83e0aeac1dc9dd7bb45e9148d2aee00779fcf32d7313fe","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"445d0e295a33c0c99d83e0aeac1dc9dd7bb45e9148d2aee00779fcf32d7313fe","first_computed_at":"2026-07-05T05:14:36.597725Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T05:14:36.597725Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"oBhrdaO88yVKNvEL7LCApE7KzGVj+6nSsN1or+A1hOcUrrHyjt80776Owaspeq/330XXTozwggKRhKhQehKIAQ==","signature_status":"signed_v1","signed_at":"2026-07-05T05:14:36.598262Z","signed_message":"canonical_sha256_bytes"},"source_id":"2209.07089","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:90cd10a8e84e879f0891fed435898055382495d48fd6345c47f172e4419c974b","sha256:aabef6f178258e5d12306a092c427bfe19012949c9a548181a05375c86c998f5"],"state_sha256":"bdca4dfe6b3f5eed9ae1bf1fb67069b642e29905c1431d84879c9559dbd1142d"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ZhTtizFjHRBQAajs6OO/e7mXqDisi8RnXX8W5E7qS44jRb0c0MWO9qTe/h1Y3Ml+pVzAhhIIyJMElA4Pmq7iDg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-09T06:01:31.122782Z","bundle_sha256":"0d633aa513ab0f51134a00aaa5d6cd88637df86ee6edfddf8a7e4163b2c06724"}}