{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:EIPZOWT6D6XVARHBWWRRDYJDRY","short_pith_number":"pith:EIPZOWT6","schema_version":"1.0","canonical_sha256":"221f975a7e1faf5044e1b5a311e1238e3abef352f0c7a8287e556c0191190cd6","source":{"kind":"arxiv","id":"1811.10996","version":1},"attestation_state":"computed","paper":{"title":"CGMH: Constrained Sentence Generation by Metropolis-Hastings Sampling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG","math.ST","stat.ML","stat.TH"],"primary_cat":"cs.CL","authors_text":"Hao Zhou, Lei Li, Lili Mou, Ning Miao, Rui Yan","submitted_at":"2018-11-14T15:46:57Z","abstract_excerpt":"In real-world applications of natural language generation, there are often constraints on the target sentences in addition to fluency and naturalness requirements. Existing language generation techniques are usually based on recurrent neural networks (RNNs). However, it is non-trivial to impose constraints on RNNs while maintaining generation quality, since RNNs generate sentences sequentially (or with beam search) from the first word to the last. In this paper, we propose CGMH, a novel approach using Metropolis-Hastings sampling for constrained sentence generation. CGMH allows complicated con"},"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":"1811.10996","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2018-11-14T15:46:57Z","cross_cats_sorted":["cs.AI","cs.LG","math.ST","stat.ML","stat.TH"],"title_canon_sha256":"6afc87ec6879413e3b179670003498e7443d3295ce2d94f5dad832a6dc889ada","abstract_canon_sha256":"e7898a20bf96f89b20c410643ea1485ae0936a5113e281e1c40180ef6e56e646"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:54:53.372006Z","signature_b64":"MtJd9KiPIzECbTKC+hdxjXyhtL3ZLA1p4K/WIDAJam96Z9InEWGZ6riU+BEJtStzascNCEQ9rzHjklykvdZcAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"221f975a7e1faf5044e1b5a311e1238e3abef352f0c7a8287e556c0191190cd6","last_reissued_at":"2026-05-17T23:54:53.371525Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:54:53.371525Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CGMH: Constrained Sentence Generation by Metropolis-Hastings Sampling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG","math.ST","stat.ML","stat.TH"],"primary_cat":"cs.CL","authors_text":"Hao Zhou, Lei Li, Lili Mou, Ning Miao, Rui Yan","submitted_at":"2018-11-14T15:46:57Z","abstract_excerpt":"In real-world applications of natural language generation, there are often constraints on the target sentences in addition to fluency and naturalness requirements. Existing language generation techniques are usually based on recurrent neural networks (RNNs). However, it is non-trivial to impose constraints on RNNs while maintaining generation quality, since RNNs generate sentences sequentially (or with beam search) from the first word to the last. In this paper, we propose CGMH, a novel approach using Metropolis-Hastings sampling for constrained sentence generation. CGMH allows complicated con"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1811.10996","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":""},"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":"1811.10996","created_at":"2026-05-17T23:54:53.371602+00:00"},{"alias_kind":"arxiv_version","alias_value":"1811.10996v1","created_at":"2026-05-17T23:54:53.371602+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1811.10996","created_at":"2026-05-17T23:54:53.371602+00:00"},{"alias_kind":"pith_short_12","alias_value":"EIPZOWT6D6XV","created_at":"2026-05-18T12:32:22.470017+00:00"},{"alias_kind":"pith_short_16","alias_value":"EIPZOWT6D6XVARHB","created_at":"2026-05-18T12:32:22.470017+00:00"},{"alias_kind":"pith_short_8","alias_value":"EIPZOWT6","created_at":"2026-05-18T12:32:22.470017+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.27039","citing_title":"Length Value Model: Scalable Value Pretraining for Token-Level Length Modeling","ref_index":15,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EIPZOWT6D6XVARHBWWRRDYJDRY","json":"https://pith.science/pith/EIPZOWT6D6XVARHBWWRRDYJDRY.json","graph_json":"https://pith.science/api/pith-number/EIPZOWT6D6XVARHBWWRRDYJDRY/graph.json","events_json":"https://pith.science/api/pith-number/EIPZOWT6D6XVARHBWWRRDYJDRY/events.json","paper":"https://pith.science/paper/EIPZOWT6"},"agent_actions":{"view_html":"https://pith.science/pith/EIPZOWT6D6XVARHBWWRRDYJDRY","download_json":"https://pith.science/pith/EIPZOWT6D6XVARHBWWRRDYJDRY.json","view_paper":"https://pith.science/paper/EIPZOWT6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1811.10996&json=true","fetch_graph":"https://pith.science/api/pith-number/EIPZOWT6D6XVARHBWWRRDYJDRY/graph.json","fetch_events":"https://pith.science/api/pith-number/EIPZOWT6D6XVARHBWWRRDYJDRY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EIPZOWT6D6XVARHBWWRRDYJDRY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EIPZOWT6D6XVARHBWWRRDYJDRY/action/storage_attestation","attest_author":"https://pith.science/pith/EIPZOWT6D6XVARHBWWRRDYJDRY/action/author_attestation","sign_citation":"https://pith.science/pith/EIPZOWT6D6XVARHBWWRRDYJDRY/action/citation_signature","submit_replication":"https://pith.science/pith/EIPZOWT6D6XVARHBWWRRDYJDRY/action/replication_record"}},"created_at":"2026-05-17T23:54:53.371602+00:00","updated_at":"2026-05-17T23:54:53.371602+00:00"}