{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:VOI3K5KN5PJYBI42VRLLOP2IFR","short_pith_number":"pith:VOI3K5KN","canonical_record":{"source":{"id":"2404.05055","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-04-07T19:29:09Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"98594d282672c8ca7f242e85b31a5aaf62a67c61fb9d50abd4447d0c92a7207e","abstract_canon_sha256":"c9e2e3f76ce055ab78d84e34c7e7cbd16bf095160cf147d79275dd84ffa62c54"},"schema_version":"1.0"},"canonical_sha256":"ab91b5754debd380a39aac56b73f482c5def125e9de132772ed50d885fe6116b","source":{"kind":"arxiv","id":"2404.05055","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2404.05055","created_at":"2026-07-05T08:05:25Z"},{"alias_kind":"arxiv_version","alias_value":"2404.05055v1","created_at":"2026-07-05T08:05:25Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.05055","created_at":"2026-07-05T08:05:25Z"},{"alias_kind":"pith_short_12","alias_value":"VOI3K5KN5PJY","created_at":"2026-07-05T08:05:25Z"},{"alias_kind":"pith_short_16","alias_value":"VOI3K5KN5PJYBI42","created_at":"2026-07-05T08:05:25Z"},{"alias_kind":"pith_short_8","alias_value":"VOI3K5KN","created_at":"2026-07-05T08:05:25Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:VOI3K5KN5PJYBI42VRLLOP2IFR","target":"record","payload":{"canonical_record":{"source":{"id":"2404.05055","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-04-07T19:29:09Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"98594d282672c8ca7f242e85b31a5aaf62a67c61fb9d50abd4447d0c92a7207e","abstract_canon_sha256":"c9e2e3f76ce055ab78d84e34c7e7cbd16bf095160cf147d79275dd84ffa62c54"},"schema_version":"1.0"},"canonical_sha256":"ab91b5754debd380a39aac56b73f482c5def125e9de132772ed50d885fe6116b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:05:25.453038Z","signature_b64":"c6mDyZoGr1OcmFC42C7tKdaEbhOcVl8zPzKAom0Lv7ROs5DBpZ96BpHTo06gsdj+rq8z59ULYzf3XlZrPP6GBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ab91b5754debd380a39aac56b73f482c5def125e9de132772ed50d885fe6116b","last_reissued_at":"2026-07-05T08:05:25.452422Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:05:25.452422Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2404.05055","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:05:25Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hPTdBXNJjUl40g1ShN7aO2qHyKRq7S3IgZFHkk8/JqO1u9Wig1tzVd5rA/SEK3EDK1YEzFlE5SIB224IoiViAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T14:52:04.661536Z"},"content_sha256":"dde590c1bb0cd9e4f8893fd1703d54f53d8201c407e2b87dd404adb86ba59287","schema_version":"1.0","event_id":"sha256:dde590c1bb0cd9e4f8893fd1703d54f53d8201c407e2b87dd404adb86ba59287"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:VOI3K5KN5PJYBI42VRLLOP2IFR","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Percentile Criterion Optimization in Offline Reinforcement Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Cyrus Cousins, Elita A. Lobo, Marek Petrik, Yair Zick","submitted_at":"2024-04-07T19:29:09Z","abstract_excerpt":"In reinforcement learning, robust policies for high-stakes decision-making problems with limited data are usually computed by optimizing the \\emph{percentile criterion}. The percentile criterion is approximately solved by constructing an \\emph{ambiguity set} that contains the true model with high probability and optimizing the policy for the worst model in the set. Since the percentile criterion is non-convex, constructing ambiguity sets is often challenging. Existing work uses \\emph{Bayesian credible regions} as ambiguity sets, but they are often unnecessarily large and result in learning ove"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.05055","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/2404.05055/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":"2026-07-05T08:05:25Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2kGfwPLosvzXlTifxpdlGNWPBv34N31ofaVK9UfjRUyV3kbOraBe0IsHFxmIFjpZdWsn2Z3KBJNTuz9xTljiBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-06T14:52:04.661913Z"},"content_sha256":"53bc9d3617d6976817bc91c4d7ab424eeef7a52f4fa5055c5bd5aaaaa26474e4","schema_version":"1.0","event_id":"sha256:53bc9d3617d6976817bc91c4d7ab424eeef7a52f4fa5055c5bd5aaaaa26474e4"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/VOI3K5KN5PJYBI42VRLLOP2IFR/bundle.json","state_url":"https://pith.science/pith/VOI3K5KN5PJYBI42VRLLOP2IFR/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/VOI3K5KN5PJYBI42VRLLOP2IFR/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":"2026-07-06T14:52:04Z","links":{"resolver":"https://pith.science/pith/VOI3K5KN5PJYBI42VRLLOP2IFR","bundle":"https://pith.science/pith/VOI3K5KN5PJYBI42VRLLOP2IFR/bundle.json","state":"https://pith.science/pith/VOI3K5KN5PJYBI42VRLLOP2IFR/state.json","well_known_bundle":"https://pith.science/.well-known/pith/VOI3K5KN5PJYBI42VRLLOP2IFR/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:VOI3K5KN5PJYBI42VRLLOP2IFR","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":"c9e2e3f76ce055ab78d84e34c7e7cbd16bf095160cf147d79275dd84ffa62c54","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-04-07T19:29:09Z","title_canon_sha256":"98594d282672c8ca7f242e85b31a5aaf62a67c61fb9d50abd4447d0c92a7207e"},"schema_version":"1.0","source":{"id":"2404.05055","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2404.05055","created_at":"2026-07-05T08:05:25Z"},{"alias_kind":"arxiv_version","alias_value":"2404.05055v1","created_at":"2026-07-05T08:05:25Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.05055","created_at":"2026-07-05T08:05:25Z"},{"alias_kind":"pith_short_12","alias_value":"VOI3K5KN5PJY","created_at":"2026-07-05T08:05:25Z"},{"alias_kind":"pith_short_16","alias_value":"VOI3K5KN5PJYBI42","created_at":"2026-07-05T08:05:25Z"},{"alias_kind":"pith_short_8","alias_value":"VOI3K5KN","created_at":"2026-07-05T08:05:25Z"}],"graph_snapshots":[{"event_id":"sha256:53bc9d3617d6976817bc91c4d7ab424eeef7a52f4fa5055c5bd5aaaaa26474e4","target":"graph","created_at":"2026-07-05T08:05:25Z","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/2404.05055/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In reinforcement learning, robust policies for high-stakes decision-making problems with limited data are usually computed by optimizing the \\emph{percentile criterion}. The percentile criterion is approximately solved by constructing an \\emph{ambiguity set} that contains the true model with high probability and optimizing the policy for the worst model in the set. Since the percentile criterion is non-convex, constructing ambiguity sets is often challenging. Existing work uses \\emph{Bayesian credible regions} as ambiguity sets, but they are often unnecessarily large and result in learning ove","authors_text":"Cyrus Cousins, Elita A. Lobo, Marek Petrik, Yair Zick","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-04-07T19:29:09Z","title":"Percentile Criterion Optimization in Offline Reinforcement Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.05055","kind":"arxiv","version":1},"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:dde590c1bb0cd9e4f8893fd1703d54f53d8201c407e2b87dd404adb86ba59287","target":"record","created_at":"2026-07-05T08:05:25Z","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":"c9e2e3f76ce055ab78d84e34c7e7cbd16bf095160cf147d79275dd84ffa62c54","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-04-07T19:29:09Z","title_canon_sha256":"98594d282672c8ca7f242e85b31a5aaf62a67c61fb9d50abd4447d0c92a7207e"},"schema_version":"1.0","source":{"id":"2404.05055","kind":"arxiv","version":1}},"canonical_sha256":"ab91b5754debd380a39aac56b73f482c5def125e9de132772ed50d885fe6116b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ab91b5754debd380a39aac56b73f482c5def125e9de132772ed50d885fe6116b","first_computed_at":"2026-07-05T08:05:25.452422Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:05:25.452422Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"c6mDyZoGr1OcmFC42C7tKdaEbhOcVl8zPzKAom0Lv7ROs5DBpZ96BpHTo06gsdj+rq8z59ULYzf3XlZrPP6GBA==","signature_status":"signed_v1","signed_at":"2026-07-05T08:05:25.453038Z","signed_message":"canonical_sha256_bytes"},"source_id":"2404.05055","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:dde590c1bb0cd9e4f8893fd1703d54f53d8201c407e2b87dd404adb86ba59287","sha256:53bc9d3617d6976817bc91c4d7ab424eeef7a52f4fa5055c5bd5aaaaa26474e4"],"state_sha256":"c6d932dd9daf68a6c89284a51e4371c95773374221b8a211d3b6694822a8d6aa"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2V13rrSZfUcDCRicFRmdNb1+0tI7dYWVn+Rmwe6m8y1QwqBuOpWAIBFud+F8PTFGlcF8VUywMQ/+V2JzMMhvAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-06T14:52:04.663926Z","bundle_sha256":"cafbf92545ea41b72d63b824e444ea48851f4131c1ef2fb6e3212a8ba03b42e1"}}