{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:7BY7EEYY4MBYRKOAFL6VRLI4JD","short_pith_number":"pith:7BY7EEYY","schema_version":"1.0","canonical_sha256":"f871f21318e30388a9c02afd58ad1c48ca6d26d943fec40e7f2ae0f39c75a2b2","source":{"kind":"arxiv","id":"2401.17426","version":1},"attestation_state":"computed","paper":{"title":"Superiority of Multi-Head Attention in In-Context Linear Regression","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jie Ren, Jiliang Tang, Pengfei He, Yingqian Cui, Yue Xing","submitted_at":"2024-01-30T20:29:06Z","abstract_excerpt":"We present a theoretical analysis of the performance of transformer with softmax attention in in-context learning with linear regression tasks. While the existing literature predominantly focuses on the convergence of transformers with single-/multi-head attention, our research centers on comparing their performance. We conduct an exact theoretical analysis to demonstrate that multi-head attention with a substantial embedding dimension performs better than single-head attention. When the number of in-context examples D increases, the prediction loss using single-/multi-head attention is in O(1"},"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":"2401.17426","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-01-30T20:29:06Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"280e4e3379c80735556a7d7c4c113c46f0abb67bb0824df59655b1c5f56b558a","abstract_canon_sha256":"09a048f6ce8942ac28011b265cc214ed177097912a122832cee37bd3464470b0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:39:36.622147Z","signature_b64":"CS818E01WjMkMj01X4C4HR/jwiQW7iMnX58Q8HjIZsV0Ul/H0zajlfWIWCVFHYqXxQp8GAoC6Vrt0A6fWRl4DQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f871f21318e30388a9c02afd58ad1c48ca6d26d943fec40e7f2ae0f39c75a2b2","last_reissued_at":"2026-07-05T07:39:36.621694Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:39:36.621694Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Superiority of Multi-Head Attention in In-Context Linear Regression","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jie Ren, Jiliang Tang, Pengfei He, Yingqian Cui, Yue Xing","submitted_at":"2024-01-30T20:29:06Z","abstract_excerpt":"We present a theoretical analysis of the performance of transformer with softmax attention in in-context learning with linear regression tasks. While the existing literature predominantly focuses on the convergence of transformers with single-/multi-head attention, our research centers on comparing their performance. We conduct an exact theoretical analysis to demonstrate that multi-head attention with a substantial embedding dimension performs better than single-head attention. When the number of in-context examples D increases, the prediction loss using single-/multi-head attention is in O(1"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.17426","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/2401.17426/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":"2401.17426","created_at":"2026-07-05T07:39:36.621750+00:00"},{"alias_kind":"arxiv_version","alias_value":"2401.17426v1","created_at":"2026-07-05T07:39:36.621750+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.17426","created_at":"2026-07-05T07:39:36.621750+00:00"},{"alias_kind":"pith_short_12","alias_value":"7BY7EEYY4MBY","created_at":"2026-07-05T07:39:36.621750+00:00"},{"alias_kind":"pith_short_16","alias_value":"7BY7EEYY4MBYRKOA","created_at":"2026-07-05T07:39:36.621750+00:00"},{"alias_kind":"pith_short_8","alias_value":"7BY7EEYY","created_at":"2026-07-05T07:39:36.621750+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.00926","citing_title":"How Transformers Learn Regular Language Recognition: A Theoretical Study on Training Dynamics and Implicit Bias","ref_index":2022,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7BY7EEYY4MBYRKOAFL6VRLI4JD","json":"https://pith.science/pith/7BY7EEYY4MBYRKOAFL6VRLI4JD.json","graph_json":"https://pith.science/api/pith-number/7BY7EEYY4MBYRKOAFL6VRLI4JD/graph.json","events_json":"https://pith.science/api/pith-number/7BY7EEYY4MBYRKOAFL6VRLI4JD/events.json","paper":"https://pith.science/paper/7BY7EEYY"},"agent_actions":{"view_html":"https://pith.science/pith/7BY7EEYY4MBYRKOAFL6VRLI4JD","download_json":"https://pith.science/pith/7BY7EEYY4MBYRKOAFL6VRLI4JD.json","view_paper":"https://pith.science/paper/7BY7EEYY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2401.17426&json=true","fetch_graph":"https://pith.science/api/pith-number/7BY7EEYY4MBYRKOAFL6VRLI4JD/graph.json","fetch_events":"https://pith.science/api/pith-number/7BY7EEYY4MBYRKOAFL6VRLI4JD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7BY7EEYY4MBYRKOAFL6VRLI4JD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7BY7EEYY4MBYRKOAFL6VRLI4JD/action/storage_attestation","attest_author":"https://pith.science/pith/7BY7EEYY4MBYRKOAFL6VRLI4JD/action/author_attestation","sign_citation":"https://pith.science/pith/7BY7EEYY4MBYRKOAFL6VRLI4JD/action/citation_signature","submit_replication":"https://pith.science/pith/7BY7EEYY4MBYRKOAFL6VRLI4JD/action/replication_record"}},"created_at":"2026-07-05T07:39:36.621750+00:00","updated_at":"2026-07-05T07:39:36.621750+00:00"}