{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:F53AASVBTE574QXMNWH4QHGLTY","short_pith_number":"pith:F53AASVB","canonical_record":{"source":{"id":"2311.02546","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-11-05T02:33:30Z","cross_cats_sorted":[],"title_canon_sha256":"13167ee8f38d9947e67c56ca9a40aa62531c02e5ed863eb553fbe979ffdab284","abstract_canon_sha256":"01cbfb06af85288048c5374213d937a75f6e88baaf696afcfb59bb1339dd2c54"},"schema_version":"1.0"},"canonical_sha256":"2f76004aa1993bfe42ec6d8fc81ccb9e3b0d24da21ce64b2d12400037a137725","source":{"kind":"arxiv","id":"2311.02546","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2311.02546","created_at":"2026-07-05T08:18:45Z"},{"alias_kind":"arxiv_version","alias_value":"2311.02546v4","created_at":"2026-07-05T08:18:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.02546","created_at":"2026-07-05T08:18:45Z"},{"alias_kind":"pith_short_12","alias_value":"F53AASVBTE57","created_at":"2026-07-05T08:18:45Z"},{"alias_kind":"pith_short_16","alias_value":"F53AASVBTE574QXM","created_at":"2026-07-05T08:18:45Z"},{"alias_kind":"pith_short_8","alias_value":"F53AASVB","created_at":"2026-07-05T08:18:45Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:F53AASVBTE574QXMNWH4QHGLTY","target":"record","payload":{"canonical_record":{"source":{"id":"2311.02546","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-11-05T02:33:30Z","cross_cats_sorted":[],"title_canon_sha256":"13167ee8f38d9947e67c56ca9a40aa62531c02e5ed863eb553fbe979ffdab284","abstract_canon_sha256":"01cbfb06af85288048c5374213d937a75f6e88baaf696afcfb59bb1339dd2c54"},"schema_version":"1.0"},"canonical_sha256":"2f76004aa1993bfe42ec6d8fc81ccb9e3b0d24da21ce64b2d12400037a137725","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:18:45.970903Z","signature_b64":"cG64IVQMH/x7nlFRVuEpFYYtTi13DYmHSLo6DqUEij+89BUWjJFM24JkDRUk6AFyH9LkysdGNGyLRwTRGqOPDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2f76004aa1993bfe42ec6d8fc81ccb9e3b0d24da21ce64b2d12400037a137725","last_reissued_at":"2026-07-05T08:18:45.970395Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:18:45.970395Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2311.02546","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:18:45Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rRaqvhBEsiZIqE4yOe8DG9h2csam++mcnkjUos4I9JVw0dep2kSQAI3nHlaFufdHRkpE+cKjIYbD4EfEDMKrBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T21:10:55.585797Z"},"content_sha256":"669951b1aafccb1e7734f7aa06e27cc44bb8e4da9f56baa00b40710a11d75d06","schema_version":"1.0","event_id":"sha256:669951b1aafccb1e7734f7aa06e27cc44bb8e4da9f56baa00b40710a11d75d06"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:F53AASVBTE574QXMNWH4QHGLTY","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"On the Second-Order Convergence of Biased Policy Gradient Algorithms","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Diego Klabjan, Siqiao Mu","submitted_at":"2023-11-05T02:33:30Z","abstract_excerpt":"Since the objective functions of reinforcement learning problems are typically highly nonconvex, it is desirable that policy gradient, the most popular algorithm, escapes saddle points and arrives at second-order stationary points. Existing results only consider vanilla policy gradient algorithms with unbiased gradient estimators, but practical implementations under the infinite-horizon discounted reward setting are biased due to finite-horizon sampling. Moreover, actor-critic methods, whose second-order convergence has not yet been established, are also biased due to the critic approximation "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.02546","kind":"arxiv","version":4},"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/2311.02546/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":"2026-07-05T08:18:45Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"TA74F+KH01RKxPkH07vKakekuB8SSuQfsTFEgyo5Cp5yxyWuAzruB+qSy3EG/gm+XazbxXzdNkMzD5XsKw7cBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T21:10:55.586693Z"},"content_sha256":"70b6c03da4450f04669d0c7e4bfc1116e961b984c2814b0fed42f2ec5cc4f4a6","schema_version":"1.0","event_id":"sha256:70b6c03da4450f04669d0c7e4bfc1116e961b984c2814b0fed42f2ec5cc4f4a6"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/F53AASVBTE574QXMNWH4QHGLTY/bundle.json","state_url":"https://pith.science/pith/F53AASVBTE574QXMNWH4QHGLTY/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/F53AASVBTE574QXMNWH4QHGLTY/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":"2026-08-05T21:10:55Z","links":{"resolver":"https://pith.science/pith/F53AASVBTE574QXMNWH4QHGLTY","bundle":"https://pith.science/pith/F53AASVBTE574QXMNWH4QHGLTY/bundle.json","state":"https://pith.science/pith/F53AASVBTE574QXMNWH4QHGLTY/state.json","well_known_bundle":"https://pith.science/.well-known/pith/F53AASVBTE574QXMNWH4QHGLTY/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:F53AASVBTE574QXMNWH4QHGLTY","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":"01cbfb06af85288048c5374213d937a75f6e88baaf696afcfb59bb1339dd2c54","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-11-05T02:33:30Z","title_canon_sha256":"13167ee8f38d9947e67c56ca9a40aa62531c02e5ed863eb553fbe979ffdab284"},"schema_version":"1.0","source":{"id":"2311.02546","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2311.02546","created_at":"2026-07-05T08:18:45Z"},{"alias_kind":"arxiv_version","alias_value":"2311.02546v4","created_at":"2026-07-05T08:18:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.02546","created_at":"2026-07-05T08:18:45Z"},{"alias_kind":"pith_short_12","alias_value":"F53AASVBTE57","created_at":"2026-07-05T08:18:45Z"},{"alias_kind":"pith_short_16","alias_value":"F53AASVBTE574QXM","created_at":"2026-07-05T08:18:45Z"},{"alias_kind":"pith_short_8","alias_value":"F53AASVB","created_at":"2026-07-05T08:18:45Z"}],"graph_snapshots":[{"event_id":"sha256:70b6c03da4450f04669d0c7e4bfc1116e961b984c2814b0fed42f2ec5cc4f4a6","target":"graph","created_at":"2026-07-05T08:18:45Z","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/2311.02546/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Since the objective functions of reinforcement learning problems are typically highly nonconvex, it is desirable that policy gradient, the most popular algorithm, escapes saddle points and arrives at second-order stationary points. Existing results only consider vanilla policy gradient algorithms with unbiased gradient estimators, but practical implementations under the infinite-horizon discounted reward setting are biased due to finite-horizon sampling. Moreover, actor-critic methods, whose second-order convergence has not yet been established, are also biased due to the critic approximation ","authors_text":"Diego Klabjan, Siqiao Mu","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-11-05T02:33:30Z","title":"On the Second-Order Convergence of Biased Policy Gradient Algorithms"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.02546","kind":"arxiv","version":4},"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":"sha256:669951b1aafccb1e7734f7aa06e27cc44bb8e4da9f56baa00b40710a11d75d06","target":"record","created_at":"2026-07-05T08:18:45Z","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":"01cbfb06af85288048c5374213d937a75f6e88baaf696afcfb59bb1339dd2c54","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-11-05T02:33:30Z","title_canon_sha256":"13167ee8f38d9947e67c56ca9a40aa62531c02e5ed863eb553fbe979ffdab284"},"schema_version":"1.0","source":{"id":"2311.02546","kind":"arxiv","version":4}},"canonical_sha256":"2f76004aa1993bfe42ec6d8fc81ccb9e3b0d24da21ce64b2d12400037a137725","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"2f76004aa1993bfe42ec6d8fc81ccb9e3b0d24da21ce64b2d12400037a137725","first_computed_at":"2026-07-05T08:18:45.970395Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:18:45.970395Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"cG64IVQMH/x7nlFRVuEpFYYtTi13DYmHSLo6DqUEij+89BUWjJFM24JkDRUk6AFyH9LkysdGNGyLRwTRGqOPDw==","signature_status":"signed_v1","signed_at":"2026-07-05T08:18:45.970903Z","signed_message":"canonical_sha256_bytes"},"source_id":"2311.02546","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:669951b1aafccb1e7734f7aa06e27cc44bb8e4da9f56baa00b40710a11d75d06","sha256:70b6c03da4450f04669d0c7e4bfc1116e961b984c2814b0fed42f2ec5cc4f4a6"],"state_sha256":"d898fb84c932202c7b84737c4199034ad3cecdae17c19b8bada1f1aad0017173"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hqpE6BvTB1wARTAb4GnfEokKfzs5y7oRQ6N/lcaTYLT7KZv5NCV/pZzMhmjpsxxv2YY1UnElW2fvHimc7ZXJCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T21:10:55.592552Z","bundle_sha256":"2fd128f185ba8f1e5115f82a286a69b9732e24f7782bbb176f2fb2975b89a4f4"}}