{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:JFU4VDSQZBMZKA36SWIHCT27IW","short_pith_number":"pith:JFU4VDSQ","schema_version":"1.0","canonical_sha256":"4969ca8e50c85995037e9590714f5f4590c689e93781e4e15feb066a66ccbac4","source":{"kind":"arxiv","id":"2505.17134","version":2},"attestation_state":"computed","paper":{"title":"LongMagpie: A Self-synthesis Method for Generating Large-scale Long-context Instructions","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Chaochen Gao, Debing Zhang, Songlin Hu, Xing Wu, Zijia Lin","submitted_at":"2025-05-22T04:05:02Z","abstract_excerpt":"High-quality long-context instruction data is essential for aligning long-context large language models (LLMs). Despite the public release of models like Qwen and Llama, their long-context instruction data remains proprietary. Human annotation is costly and challenging, while template-based synthesis methods limit scale, diversity, and quality. We introduce LongMagpie, a self-synthesis framework that automatically generates large-scale long-context instruction data. Our key insight is that aligned long-context LLMs, when presented with a document followed by special tokens preceding a user tur"},"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":"2505.17134","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-05-22T04:05:02Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"5bb3b4494685ad0993d9d26cfe465ae9f417fc00a6894dc8bb2f6893dba970de","abstract_canon_sha256":"677088529abaf44954619f8959da48a9f928dcf12284d8ed712906cf45b759e6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:14:29.259388Z","signature_b64":"jeklnLk8n5eULVpg4/g66O4LwYENZn+W+tJXTR2LdU0EPc7sqYX5vRVo1T8OuPEBU98aQvrR4fAGe7vgcXMNAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4969ca8e50c85995037e9590714f5f4590c689e93781e4e15feb066a66ccbac4","last_reissued_at":"2026-07-05T11:14:29.258882Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:14:29.258882Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LongMagpie: A Self-synthesis Method for Generating Large-scale Long-context Instructions","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Chaochen Gao, Debing Zhang, Songlin Hu, Xing Wu, Zijia Lin","submitted_at":"2025-05-22T04:05:02Z","abstract_excerpt":"High-quality long-context instruction data is essential for aligning long-context large language models (LLMs). Despite the public release of models like Qwen and Llama, their long-context instruction data remains proprietary. Human annotation is costly and challenging, while template-based synthesis methods limit scale, diversity, and quality. We introduce LongMagpie, a self-synthesis framework that automatically generates large-scale long-context instruction data. Our key insight is that aligned long-context LLMs, when presented with a document followed by special tokens preceding a user tur"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.17134","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/2505.17134/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":"2505.17134","created_at":"2026-07-05T11:14:29.258947+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.17134v2","created_at":"2026-07-05T11:14:29.258947+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.17134","created_at":"2026-07-05T11:14:29.258947+00:00"},{"alias_kind":"pith_short_12","alias_value":"JFU4VDSQZBMZ","created_at":"2026-07-05T11:14:29.258947+00:00"},{"alias_kind":"pith_short_16","alias_value":"JFU4VDSQZBMZKA36","created_at":"2026-07-05T11:14:29.258947+00:00"},{"alias_kind":"pith_short_8","alias_value":"JFU4VDSQ","created_at":"2026-07-05T11:14:29.258947+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.07809","citing_title":"PolicyLong: Towards On-Policy Context Extension","ref_index":4,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JFU4VDSQZBMZKA36SWIHCT27IW","json":"https://pith.science/pith/JFU4VDSQZBMZKA36SWIHCT27IW.json","graph_json":"https://pith.science/api/pith-number/JFU4VDSQZBMZKA36SWIHCT27IW/graph.json","events_json":"https://pith.science/api/pith-number/JFU4VDSQZBMZKA36SWIHCT27IW/events.json","paper":"https://pith.science/paper/JFU4VDSQ"},"agent_actions":{"view_html":"https://pith.science/pith/JFU4VDSQZBMZKA36SWIHCT27IW","download_json":"https://pith.science/pith/JFU4VDSQZBMZKA36SWIHCT27IW.json","view_paper":"https://pith.science/paper/JFU4VDSQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.17134&json=true","fetch_graph":"https://pith.science/api/pith-number/JFU4VDSQZBMZKA36SWIHCT27IW/graph.json","fetch_events":"https://pith.science/api/pith-number/JFU4VDSQZBMZKA36SWIHCT27IW/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JFU4VDSQZBMZKA36SWIHCT27IW/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JFU4VDSQZBMZKA36SWIHCT27IW/action/storage_attestation","attest_author":"https://pith.science/pith/JFU4VDSQZBMZKA36SWIHCT27IW/action/author_attestation","sign_citation":"https://pith.science/pith/JFU4VDSQZBMZKA36SWIHCT27IW/action/citation_signature","submit_replication":"https://pith.science/pith/JFU4VDSQZBMZKA36SWIHCT27IW/action/replication_record"}},"created_at":"2026-07-05T11:14:29.258947+00:00","updated_at":"2026-07-05T11:14:29.258947+00:00"}