{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:TJZ7KLWVMQ74LWSJLMZMDVWTTL","short_pith_number":"pith:TJZ7KLWV","schema_version":"1.0","canonical_sha256":"9a73f52ed5643fc5da495b32c1d6d39af3c7bb13cf14553875caa16979190005","source":{"kind":"arxiv","id":"2409.12181","version":2},"attestation_state":"computed","paper":{"title":"A Controlled Study on Long Context Extension and Generalization in LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Alexander M. Rush, Fei Yuan, Jing Nathan Yan, Justin T. Chiu, Siyu Ren, Songlin Yang, Wenting Zhao, Yi Lu, Zhiyong Wu","submitted_at":"2024-09-18T17:53:17Z","abstract_excerpt":"Broad textual understanding and in-context learning require language models that utilize full document contexts. Due to the implementation challenges associated with directly training long-context models, many methods have been proposed for extending models to handle long contexts. However, owing to differences in data and model classes, it has been challenging to compare these approaches, leading to uncertainty as to how to evaluate long-context performance and whether it differs from standard evaluation. We implement a controlled protocol for extension methods with a standardized evaluation,"},"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":"2409.12181","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-18T17:53:17Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"d8310641771eabf0b3779f8c1f6a1f7c8f263407e322c0915815fdf8773d8a97","abstract_canon_sha256":"ecb9a12f9e91d5851b02e20c464904d4c55c72cfa8dcb0e6540db356c295dc4b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:10:32.773699Z","signature_b64":"MYW7cXvEasTWYLQoi7plQoR7q69c0zfRJNxPSqhHWY0XbYg49PADZ7b+OZBM42c+pEy/WApfWPrJ4op4D0ZABA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9a73f52ed5643fc5da495b32c1d6d39af3c7bb13cf14553875caa16979190005","last_reissued_at":"2026-07-05T09:10:32.773267Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:10:32.773267Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"A Controlled Study on Long Context Extension and Generalization in LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Alexander M. Rush, Fei Yuan, Jing Nathan Yan, Justin T. Chiu, Siyu Ren, Songlin Yang, Wenting Zhao, Yi Lu, Zhiyong Wu","submitted_at":"2024-09-18T17:53:17Z","abstract_excerpt":"Broad textual understanding and in-context learning require language models that utilize full document contexts. Due to the implementation challenges associated with directly training long-context models, many methods have been proposed for extending models to handle long contexts. However, owing to differences in data and model classes, it has been challenging to compare these approaches, leading to uncertainty as to how to evaluate long-context performance and whether it differs from standard evaluation. We implement a controlled protocol for extension methods with a standardized evaluation,"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.12181","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/2409.12181/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":"2409.12181","created_at":"2026-07-05T09:10:32.773325+00:00"},{"alias_kind":"arxiv_version","alias_value":"2409.12181v2","created_at":"2026-07-05T09:10:32.773325+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.12181","created_at":"2026-07-05T09:10:32.773325+00:00"},{"alias_kind":"pith_short_12","alias_value":"TJZ7KLWVMQ74","created_at":"2026-07-05T09:10:32.773325+00:00"},{"alias_kind":"pith_short_16","alias_value":"TJZ7KLWVMQ74LWSJ","created_at":"2026-07-05T09:10:32.773325+00:00"},{"alias_kind":"pith_short_8","alias_value":"TJZ7KLWV","created_at":"2026-07-05T09:10:32.773325+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.10963","citing_title":"AROMA: Mixed-Initiative AI Assistance for Non-Visual Cooking by Grounding Multi-modal Information Between Reality and Videos","ref_index":54,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TJZ7KLWVMQ74LWSJLMZMDVWTTL","json":"https://pith.science/pith/TJZ7KLWVMQ74LWSJLMZMDVWTTL.json","graph_json":"https://pith.science/api/pith-number/TJZ7KLWVMQ74LWSJLMZMDVWTTL/graph.json","events_json":"https://pith.science/api/pith-number/TJZ7KLWVMQ74LWSJLMZMDVWTTL/events.json","paper":"https://pith.science/paper/TJZ7KLWV"},"agent_actions":{"view_html":"https://pith.science/pith/TJZ7KLWVMQ74LWSJLMZMDVWTTL","download_json":"https://pith.science/pith/TJZ7KLWVMQ74LWSJLMZMDVWTTL.json","view_paper":"https://pith.science/paper/TJZ7KLWV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2409.12181&json=true","fetch_graph":"https://pith.science/api/pith-number/TJZ7KLWVMQ74LWSJLMZMDVWTTL/graph.json","fetch_events":"https://pith.science/api/pith-number/TJZ7KLWVMQ74LWSJLMZMDVWTTL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TJZ7KLWVMQ74LWSJLMZMDVWTTL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TJZ7KLWVMQ74LWSJLMZMDVWTTL/action/storage_attestation","attest_author":"https://pith.science/pith/TJZ7KLWVMQ74LWSJLMZMDVWTTL/action/author_attestation","sign_citation":"https://pith.science/pith/TJZ7KLWVMQ74LWSJLMZMDVWTTL/action/citation_signature","submit_replication":"https://pith.science/pith/TJZ7KLWVMQ74LWSJLMZMDVWTTL/action/replication_record"}},"created_at":"2026-07-05T09:10:32.773325+00:00","updated_at":"2026-07-05T09:10:32.773325+00:00"}