{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:UEAXDXOCJAZACYPNZDD6XEPAXZ","short_pith_number":"pith:UEAXDXOC","schema_version":"1.0","canonical_sha256":"a10171ddc248320161edc8c7eb91e0be541bb3c8d22057836b2d0ef32cdf480f","source":{"kind":"arxiv","id":"2401.18058","version":1},"attestation_state":"computed","paper":{"title":"LongAlign: A Recipe for Long Context Alignment of Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Jiajie Zhang, Jie Tang, Ji Qi, Juanzi Li, Lei Hou, Xin Lv, Yushi Bai, Yuxiao Dong, Yuze He","submitted_at":"2024-01-31T18:29:39Z","abstract_excerpt":"Extending large language models to effectively handle long contexts requires instruction fine-tuning on input sequences of similar length. To address this, we present LongAlign -- a recipe of the instruction data, training, and evaluation for long context alignment. First, we construct a long instruction-following dataset using Self-Instruct. To ensure the data diversity, it covers a broad range of tasks from various long context sources. Second, we adopt the packing and sorted batching strategies to speed up supervised fine-tuning on data with varied length distributions. Additionally, we dev"},"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":"2401.18058","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-01-31T18:29:39Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"5b8277ad8892b63026a65847f024a9c993f4d49e5e5f8aedc8797a873f1bbfa3","abstract_canon_sha256":"af53762249a6050e030e50b037ffab5f10250ffd8fa34f1f6699c10fc5365d81"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:39:46.305357Z","signature_b64":"5tgL/4dEcotjXY/qjCSFPV428EHHOcfdIoVp/eNTtRKfczxN5c+icaMIaQtBu0bXSu5UJVvgmFu2WcuAHaNJBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a10171ddc248320161edc8c7eb91e0be541bb3c8d22057836b2d0ef32cdf480f","last_reissued_at":"2026-07-05T07:39:46.304933Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:39:46.304933Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LongAlign: A Recipe for Long Context Alignment of Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Jiajie Zhang, Jie Tang, Ji Qi, Juanzi Li, Lei Hou, Xin Lv, Yushi Bai, Yuxiao Dong, Yuze He","submitted_at":"2024-01-31T18:29:39Z","abstract_excerpt":"Extending large language models to effectively handle long contexts requires instruction fine-tuning on input sequences of similar length. To address this, we present LongAlign -- a recipe of the instruction data, training, and evaluation for long context alignment. First, we construct a long instruction-following dataset using Self-Instruct. To ensure the data diversity, it covers a broad range of tasks from various long context sources. Second, we adopt the packing and sorted batching strategies to speed up supervised fine-tuning on data with varied length distributions. Additionally, we dev"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.18058","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/2401.18058/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":"2401.18058","created_at":"2026-07-05T07:39:46.304988+00:00"},{"alias_kind":"arxiv_version","alias_value":"2401.18058v1","created_at":"2026-07-05T07:39:46.304988+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.18058","created_at":"2026-07-05T07:39:46.304988+00:00"},{"alias_kind":"pith_short_12","alias_value":"UEAXDXOCJAZA","created_at":"2026-07-05T07:39:46.304988+00:00"},{"alias_kind":"pith_short_16","alias_value":"UEAXDXOCJAZACYPN","created_at":"2026-07-05T07:39:46.304988+00:00"},{"alias_kind":"pith_short_8","alias_value":"UEAXDXOC","created_at":"2026-07-05T07:39:46.304988+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":8,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.14589","citing_title":"EndPrompt: Efficient Long-Context Extension via Terminal Anchoring","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2605.14589","citing_title":"EndPrompt: Efficient Long-Context Extension via Terminal Anchoring","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2502.02737","citing_title":"SmolLM2: When Smol Goes Big -- Data-Centric Training of a Small Language Model","ref_index":150,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10544","citing_title":"Where Does Long-Context Supervision Actually Go? Effective-Context Exposure Balancing","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18258","citing_title":"Long-Text-to-Image Generation via Compositional Prompt Decomposition","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2604.07981","citing_title":"A Decomposition Perspective to Long-context Reasoning for LLMs","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2604.05151","citing_title":"Context Collapse: Barriers to Adoption for Generative AI in Workplace Settings","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2408.06072","citing_title":"CogVideoX: Text-to-Video Diffusion Models with An Expert Transformer","ref_index":12,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UEAXDXOCJAZACYPNZDD6XEPAXZ","json":"https://pith.science/pith/UEAXDXOCJAZACYPNZDD6XEPAXZ.json","graph_json":"https://pith.science/api/pith-number/UEAXDXOCJAZACYPNZDD6XEPAXZ/graph.json","events_json":"https://pith.science/api/pith-number/UEAXDXOCJAZACYPNZDD6XEPAXZ/events.json","paper":"https://pith.science/paper/UEAXDXOC"},"agent_actions":{"view_html":"https://pith.science/pith/UEAXDXOCJAZACYPNZDD6XEPAXZ","download_json":"https://pith.science/pith/UEAXDXOCJAZACYPNZDD6XEPAXZ.json","view_paper":"https://pith.science/paper/UEAXDXOC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2401.18058&json=true","fetch_graph":"https://pith.science/api/pith-number/UEAXDXOCJAZACYPNZDD6XEPAXZ/graph.json","fetch_events":"https://pith.science/api/pith-number/UEAXDXOCJAZACYPNZDD6XEPAXZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UEAXDXOCJAZACYPNZDD6XEPAXZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UEAXDXOCJAZACYPNZDD6XEPAXZ/action/storage_attestation","attest_author":"https://pith.science/pith/UEAXDXOCJAZACYPNZDD6XEPAXZ/action/author_attestation","sign_citation":"https://pith.science/pith/UEAXDXOCJAZACYPNZDD6XEPAXZ/action/citation_signature","submit_replication":"https://pith.science/pith/UEAXDXOCJAZACYPNZDD6XEPAXZ/action/replication_record"}},"created_at":"2026-07-05T07:39:46.304988+00:00","updated_at":"2026-07-05T07:39:46.304988+00:00"}