{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:7IKQDUZ7CUX3JVRTJYDEP7DVRO","short_pith_number":"pith:7IKQDUZ7","schema_version":"1.0","canonical_sha256":"fa1501d33f152fb4d6334e0647fc758baca42aef3399eff2bf0154de6eb975a6","source":{"kind":"arxiv","id":"1912.13151","version":2},"attestation_state":"computed","paper":{"title":"Adaptive Correlated Monte Carlo for Contextual Categorical Sequence Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Mingyuan Zhou, Xinjie Fan, Yizhe Zhang, Zhendong Wang","submitted_at":"2019-12-31T03:01:55Z","abstract_excerpt":"Sequence generation models are commonly refined with reinforcement learning over user-defined metrics. However, high gradient variance hinders the practical use of this method. To stabilize this method, we adapt to contextual generation of categorical sequences a policy gradient estimator, which evaluates a set of correlated Monte Carlo (MC) rollouts for variance control. Due to the correlation, the number of unique rollouts is random and adaptive to model uncertainty; those rollouts naturally become baselines for each other, and hence are combined to effectively reduce gradient variance. We a"},"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":"1912.13151","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2019-12-31T03:01:55Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"e8661c24375d88f99a0ea61eed8d32594a01e59d8f7315c77badfe3e1d9350a7","abstract_canon_sha256":"707cc79d38e01a14fd5a0a758b43424ff07163deb3858ac4f2896ccc61840a57"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:10:57.209959Z","signature_b64":"LWq4pT3bhWFgbO7K5a3KQ0r5TJ9rCYsPR01NDYIAqIBy99zZpMVpRXiMimHaOojd7vrElfDUCBU3peI+bF1vBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fa1501d33f152fb4d6334e0647fc758baca42aef3399eff2bf0154de6eb975a6","last_reissued_at":"2026-07-05T01:10:57.209498Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:10:57.209498Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Adaptive Correlated Monte Carlo for Contextual Categorical Sequence Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Mingyuan Zhou, Xinjie Fan, Yizhe Zhang, Zhendong Wang","submitted_at":"2019-12-31T03:01:55Z","abstract_excerpt":"Sequence generation models are commonly refined with reinforcement learning over user-defined metrics. However, high gradient variance hinders the practical use of this method. To stabilize this method, we adapt to contextual generation of categorical sequences a policy gradient estimator, which evaluates a set of correlated Monte Carlo (MC) rollouts for variance control. Due to the correlation, the number of unique rollouts is random and adaptive to model uncertainty; those rollouts naturally become baselines for each other, and hence are combined to effectively reduce gradient variance. We a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1912.13151","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/1912.13151/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":"1912.13151","created_at":"2026-07-05T01:10:57.209558+00:00"},{"alias_kind":"arxiv_version","alias_value":"1912.13151v2","created_at":"2026-07-05T01:10:57.209558+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1912.13151","created_at":"2026-07-05T01:10:57.209558+00:00"},{"alias_kind":"pith_short_12","alias_value":"7IKQDUZ7CUX3","created_at":"2026-07-05T01:10:57.209558+00:00"},{"alias_kind":"pith_short_16","alias_value":"7IKQDUZ7CUX3JVRT","created_at":"2026-07-05T01:10:57.209558+00:00"},{"alias_kind":"pith_short_8","alias_value":"7IKQDUZ7","created_at":"2026-07-05T01:10:57.209558+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7IKQDUZ7CUX3JVRTJYDEP7DVRO","json":"https://pith.science/pith/7IKQDUZ7CUX3JVRTJYDEP7DVRO.json","graph_json":"https://pith.science/api/pith-number/7IKQDUZ7CUX3JVRTJYDEP7DVRO/graph.json","events_json":"https://pith.science/api/pith-number/7IKQDUZ7CUX3JVRTJYDEP7DVRO/events.json","paper":"https://pith.science/paper/7IKQDUZ7"},"agent_actions":{"view_html":"https://pith.science/pith/7IKQDUZ7CUX3JVRTJYDEP7DVRO","download_json":"https://pith.science/pith/7IKQDUZ7CUX3JVRTJYDEP7DVRO.json","view_paper":"https://pith.science/paper/7IKQDUZ7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1912.13151&json=true","fetch_graph":"https://pith.science/api/pith-number/7IKQDUZ7CUX3JVRTJYDEP7DVRO/graph.json","fetch_events":"https://pith.science/api/pith-number/7IKQDUZ7CUX3JVRTJYDEP7DVRO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7IKQDUZ7CUX3JVRTJYDEP7DVRO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7IKQDUZ7CUX3JVRTJYDEP7DVRO/action/storage_attestation","attest_author":"https://pith.science/pith/7IKQDUZ7CUX3JVRTJYDEP7DVRO/action/author_attestation","sign_citation":"https://pith.science/pith/7IKQDUZ7CUX3JVRTJYDEP7DVRO/action/citation_signature","submit_replication":"https://pith.science/pith/7IKQDUZ7CUX3JVRTJYDEP7DVRO/action/replication_record"}},"created_at":"2026-07-05T01:10:57.209558+00:00","updated_at":"2026-07-05T01:10:57.209558+00:00"}