{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:TFCLLAZFWT6BG6DNSUPNW62RDO","short_pith_number":"pith:TFCLLAZF","schema_version":"1.0","canonical_sha256":"9944b58325b4fc13786d951edb7b511b995d49755c2cd81fd0095e1fa23fc1c2","source":{"kind":"arxiv","id":"1902.01370","version":3},"attestation_state":"computed","paper":{"title":"Insertion-based Decoding with automatically Inferred Generation Order","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Jiatao Gu, Kyunghyun Cho, Qi Liu","submitted_at":"2019-02-04T18:35:59Z","abstract_excerpt":"Conventional neural autoregressive decoding commonly assumes a fixed left-to-right generation order, which may be sub-optimal. In this work, we propose a novel decoding algorithm -- InDIGO -- which supports flexible sequence generation in arbitrary orders through insertion operations. We extend Transformer, a state-of-the-art sequence generation model, to efficiently implement the proposed approach, enabling it to be trained with either a pre-defined generation order or adaptive orders obtained from beam-search. Experiments on four real-world tasks, including word order recovery, machine trans"},"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":"1902.01370","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2019-02-04T18:35:59Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"c94ed5f01762a05f0c8cd6da889287552da4eb5b805db1fd2962558f0951dbcc","abstract_canon_sha256":"28bbe95cc2faacc26caa311dd0dd7a55b9d4f2edd32e42b310d1f65a0ea24930"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:15:02.705375Z","signature_b64":"DvVtIQd9QYnjwshLftrxOLhlE/aqBLpWNtkbJJYfbKt6tbKH0DvXxkcxdoNx32dfNUnSGyHv5mJzomf808R5AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9944b58325b4fc13786d951edb7b511b995d49755c2cd81fd0095e1fa23fc1c2","last_reissued_at":"2026-07-05T00:15:02.704908Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:15:02.704908Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Insertion-based Decoding with automatically Inferred Generation Order","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Jiatao Gu, Kyunghyun Cho, Qi Liu","submitted_at":"2019-02-04T18:35:59Z","abstract_excerpt":"Conventional neural autoregressive decoding commonly assumes a fixed left-to-right generation order, which may be sub-optimal. In this work, we propose a novel decoding algorithm -- InDIGO -- which supports flexible sequence generation in arbitrary orders through insertion operations. We extend Transformer, a state-of-the-art sequence generation model, to efficiently implement the proposed approach, enabling it to be trained with either a pre-defined generation order or adaptive orders obtained from beam-search. Experiments on four real-world tasks, including word order recovery, machine trans"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1902.01370","kind":"arxiv","version":3},"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/1902.01370/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":"1902.01370","created_at":"2026-07-05T00:15:02.704963+00:00"},{"alias_kind":"arxiv_version","alias_value":"1902.01370v3","created_at":"2026-07-05T00:15:02.704963+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1902.01370","created_at":"2026-07-05T00:15:02.704963+00:00"},{"alias_kind":"pith_short_12","alias_value":"TFCLLAZFWT6B","created_at":"2026-07-05T00:15:02.704963+00:00"},{"alias_kind":"pith_short_16","alias_value":"TFCLLAZFWT6BG6DN","created_at":"2026-07-05T00:15:02.704963+00:00"},{"alias_kind":"pith_short_8","alias_value":"TFCLLAZF","created_at":"2026-07-05T00:15:02.704963+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"1909.02480","citing_title":"FlowSeq: Non-Autoregressive Conditional Sequence Generation with Generative Flow","ref_index":11,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TFCLLAZFWT6BG6DNSUPNW62RDO","json":"https://pith.science/pith/TFCLLAZFWT6BG6DNSUPNW62RDO.json","graph_json":"https://pith.science/api/pith-number/TFCLLAZFWT6BG6DNSUPNW62RDO/graph.json","events_json":"https://pith.science/api/pith-number/TFCLLAZFWT6BG6DNSUPNW62RDO/events.json","paper":"https://pith.science/paper/TFCLLAZF"},"agent_actions":{"view_html":"https://pith.science/pith/TFCLLAZFWT6BG6DNSUPNW62RDO","download_json":"https://pith.science/pith/TFCLLAZFWT6BG6DNSUPNW62RDO.json","view_paper":"https://pith.science/paper/TFCLLAZF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1902.01370&json=true","fetch_graph":"https://pith.science/api/pith-number/TFCLLAZFWT6BG6DNSUPNW62RDO/graph.json","fetch_events":"https://pith.science/api/pith-number/TFCLLAZFWT6BG6DNSUPNW62RDO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TFCLLAZFWT6BG6DNSUPNW62RDO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TFCLLAZFWT6BG6DNSUPNW62RDO/action/storage_attestation","attest_author":"https://pith.science/pith/TFCLLAZFWT6BG6DNSUPNW62RDO/action/author_attestation","sign_citation":"https://pith.science/pith/TFCLLAZFWT6BG6DNSUPNW62RDO/action/citation_signature","submit_replication":"https://pith.science/pith/TFCLLAZFWT6BG6DNSUPNW62RDO/action/replication_record"}},"created_at":"2026-07-05T00:15:02.704963+00:00","updated_at":"2026-07-05T00:15:02.704963+00:00"}