{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:4C7XPL7EK6IH3Z3YSQP5S75BXT","short_pith_number":"pith:4C7XPL7E","schema_version":"1.0","canonical_sha256":"e0bf77afe457907de778941fd97fa1bcd29f032ec723642e56d0566eaf35bc7d","source":{"kind":"arxiv","id":"2506.09251","version":2},"attestation_state":"computed","paper":{"title":"Extrapolation by Association: Length Generalization Transfer in Transformers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Avi Schwarzschild, Dimitris Papailiopoulos, Nayoung Lee, Samet Oymak, Ziyang Cai","submitted_at":"2025-06-10T21:22:51Z","abstract_excerpt":"Transformer language models have demonstrated impressive generalization capabilities in natural language domains, yet we lack a fine-grained understanding of how such generalization arises. In this paper, we investigate length generalization--the ability to extrapolate from shorter to longer inputs--through the lens of \\textit{task association}. We find that length generalization can be \\textit{transferred} across related tasks. That is, training a model with a longer and related auxiliary task can lead it to generalize to unseen and longer inputs from some other target task. We demonstrate th"},"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":"2506.09251","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-10T21:22:51Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"293f8ecd1ae84c554660674241211af87341f4f114810365b9ac62bfc525ed37","abstract_canon_sha256":"b19d6ef2b29c0e7ab38c1b27935366402721de61ec89108ee86375d2a61cd751"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:48:04.537788Z","signature_b64":"lk+GRd3UYNAYj1mk1KMXzB48G0QKvCy/n/Bzrfu7rFrHM8gumnPGDRmkto1Mbs3ZLfp8OXBlPVdUhiG7EQBFDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e0bf77afe457907de778941fd97fa1bcd29f032ec723642e56d0566eaf35bc7d","last_reissued_at":"2026-07-05T11:48:04.537280Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:48:04.537280Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Extrapolation by Association: Length Generalization Transfer in Transformers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Avi Schwarzschild, Dimitris Papailiopoulos, Nayoung Lee, Samet Oymak, Ziyang Cai","submitted_at":"2025-06-10T21:22:51Z","abstract_excerpt":"Transformer language models have demonstrated impressive generalization capabilities in natural language domains, yet we lack a fine-grained understanding of how such generalization arises. In this paper, we investigate length generalization--the ability to extrapolate from shorter to longer inputs--through the lens of \\textit{task association}. We find that length generalization can be \\textit{transferred} across related tasks. That is, training a model with a longer and related auxiliary task can lead it to generalize to unseen and longer inputs from some other target task. We demonstrate th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.09251","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/2506.09251/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":"2506.09251","created_at":"2026-07-05T11:48:04.537340+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.09251v2","created_at":"2026-07-05T11:48:04.537340+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.09251","created_at":"2026-07-05T11:48:04.537340+00:00"},{"alias_kind":"pith_short_12","alias_value":"4C7XPL7EK6IH","created_at":"2026-07-05T11:48:04.537340+00:00"},{"alias_kind":"pith_short_16","alias_value":"4C7XPL7EK6IH3Z3Y","created_at":"2026-07-05T11:48:04.537340+00:00"},{"alias_kind":"pith_short_8","alias_value":"4C7XPL7E","created_at":"2026-07-05T11:48:04.537340+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.29983","citing_title":"Stabilizing Extrapolation in Looped Transformers via Learned Stochastic Stopping","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15306","citing_title":"Generalization in LLM Problem Solving: The Case of the Shortest Path","ref_index":7,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4C7XPL7EK6IH3Z3YSQP5S75BXT","json":"https://pith.science/pith/4C7XPL7EK6IH3Z3YSQP5S75BXT.json","graph_json":"https://pith.science/api/pith-number/4C7XPL7EK6IH3Z3YSQP5S75BXT/graph.json","events_json":"https://pith.science/api/pith-number/4C7XPL7EK6IH3Z3YSQP5S75BXT/events.json","paper":"https://pith.science/paper/4C7XPL7E"},"agent_actions":{"view_html":"https://pith.science/pith/4C7XPL7EK6IH3Z3YSQP5S75BXT","download_json":"https://pith.science/pith/4C7XPL7EK6IH3Z3YSQP5S75BXT.json","view_paper":"https://pith.science/paper/4C7XPL7E","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.09251&json=true","fetch_graph":"https://pith.science/api/pith-number/4C7XPL7EK6IH3Z3YSQP5S75BXT/graph.json","fetch_events":"https://pith.science/api/pith-number/4C7XPL7EK6IH3Z3YSQP5S75BXT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4C7XPL7EK6IH3Z3YSQP5S75BXT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4C7XPL7EK6IH3Z3YSQP5S75BXT/action/storage_attestation","attest_author":"https://pith.science/pith/4C7XPL7EK6IH3Z3YSQP5S75BXT/action/author_attestation","sign_citation":"https://pith.science/pith/4C7XPL7EK6IH3Z3YSQP5S75BXT/action/citation_signature","submit_replication":"https://pith.science/pith/4C7XPL7EK6IH3Z3YSQP5S75BXT/action/replication_record"}},"created_at":"2026-07-05T11:48:04.537340+00:00","updated_at":"2026-07-05T11:48:04.537340+00:00"}