{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:CTJWY2ANDFQJYZDEQHHTDLD3ZT","short_pith_number":"pith:CTJWY2AN","schema_version":"1.0","canonical_sha256":"14d36c680d19609c646481cf31ac7bccd5f9636affba1a915f9885bd406a62d6","source":{"kind":"arxiv","id":"2502.12304","version":1},"attestation_state":"computed","paper":{"title":"Warmup Generations: A Task-Agnostic Approach for Guiding Sequence-to-Sequence Learning with Unsupervised Initial State Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"David Ifeoluwa Adelani, Jiayi Wang, Pontus Stenetorp, Senyu Li, Siva Reddy, Xue Liu, Zipeng Sun","submitted_at":"2025-02-17T20:23:42Z","abstract_excerpt":"Traditional supervised fine-tuning (SFT) strategies for sequence-to-sequence tasks often train models to directly generate the target output. Recent work has shown that guiding models with intermediate steps, such as keywords, outlines, or reasoning chains, can significantly improve performance, coherence, and interpretability. However, these methods often depend on predefined intermediate formats and annotated data, limiting their scalability and generalizability. In this work, we introduce a task-agnostic framework that enables models to generate intermediate \"warmup\" sequences. These warmup"},"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":"2502.12304","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-17T20:23:42Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"c2e6542129dae32a03a93e79614c239a506faacb45878a09260b63a16dd51e26","abstract_canon_sha256":"fe602ab38284c7808e40acf79c53b406880a7d3aa5865980b2ce9ae18d770ac3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:15:53.313503Z","signature_b64":"RCLuq6736MmboGGKI2hk6KjCEEdN7bOMxNd8NA3ji5febZtjtm/ZYuUe3Y3KVVmtpPZiE5ni3Xs1zmdmipUUDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"14d36c680d19609c646481cf31ac7bccd5f9636affba1a915f9885bd406a62d6","last_reissued_at":"2026-07-05T10:15:53.312983Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:15:53.312983Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Warmup Generations: A Task-Agnostic Approach for Guiding Sequence-to-Sequence Learning with Unsupervised Initial State Generation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"David Ifeoluwa Adelani, Jiayi Wang, Pontus Stenetorp, Senyu Li, Siva Reddy, Xue Liu, Zipeng Sun","submitted_at":"2025-02-17T20:23:42Z","abstract_excerpt":"Traditional supervised fine-tuning (SFT) strategies for sequence-to-sequence tasks often train models to directly generate the target output. Recent work has shown that guiding models with intermediate steps, such as keywords, outlines, or reasoning chains, can significantly improve performance, coherence, and interpretability. However, these methods often depend on predefined intermediate formats and annotated data, limiting their scalability and generalizability. In this work, we introduce a task-agnostic framework that enables models to generate intermediate \"warmup\" sequences. These warmup"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.12304","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/2502.12304/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":"2502.12304","created_at":"2026-07-05T10:15:53.313044+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.12304v1","created_at":"2026-07-05T10:15:53.313044+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.12304","created_at":"2026-07-05T10:15:53.313044+00:00"},{"alias_kind":"pith_short_12","alias_value":"CTJWY2ANDFQJ","created_at":"2026-07-05T10:15:53.313044+00:00"},{"alias_kind":"pith_short_16","alias_value":"CTJWY2ANDFQJYZDE","created_at":"2026-07-05T10:15:53.313044+00:00"},{"alias_kind":"pith_short_8","alias_value":"CTJWY2AN","created_at":"2026-07-05T10:15:53.313044+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/CTJWY2ANDFQJYZDEQHHTDLD3ZT","json":"https://pith.science/pith/CTJWY2ANDFQJYZDEQHHTDLD3ZT.json","graph_json":"https://pith.science/api/pith-number/CTJWY2ANDFQJYZDEQHHTDLD3ZT/graph.json","events_json":"https://pith.science/api/pith-number/CTJWY2ANDFQJYZDEQHHTDLD3ZT/events.json","paper":"https://pith.science/paper/CTJWY2AN"},"agent_actions":{"view_html":"https://pith.science/pith/CTJWY2ANDFQJYZDEQHHTDLD3ZT","download_json":"https://pith.science/pith/CTJWY2ANDFQJYZDEQHHTDLD3ZT.json","view_paper":"https://pith.science/paper/CTJWY2AN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.12304&json=true","fetch_graph":"https://pith.science/api/pith-number/CTJWY2ANDFQJYZDEQHHTDLD3ZT/graph.json","fetch_events":"https://pith.science/api/pith-number/CTJWY2ANDFQJYZDEQHHTDLD3ZT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CTJWY2ANDFQJYZDEQHHTDLD3ZT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CTJWY2ANDFQJYZDEQHHTDLD3ZT/action/storage_attestation","attest_author":"https://pith.science/pith/CTJWY2ANDFQJYZDEQHHTDLD3ZT/action/author_attestation","sign_citation":"https://pith.science/pith/CTJWY2ANDFQJYZDEQHHTDLD3ZT/action/citation_signature","submit_replication":"https://pith.science/pith/CTJWY2ANDFQJYZDEQHHTDLD3ZT/action/replication_record"}},"created_at":"2026-07-05T10:15:53.313044+00:00","updated_at":"2026-07-05T10:15:53.313044+00:00"}