{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:EXT7HVM5YVNKH25BGG36BYIAXM","short_pith_number":"pith:EXT7HVM5","schema_version":"1.0","canonical_sha256":"25e7f3d59dc55aa3eba131b7e0e100bb3bd3ba16c6657bd1216915d8b6b5ba31","source":{"kind":"arxiv","id":"2307.14023","version":3},"attestation_state":"computed","paper":{"title":"Are Transformers with One Layer Self-Attention Using Low-Rank Weight Matrices Universal Approximators?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Issei Sato, Tokio Kajitsuka","submitted_at":"2023-07-26T08:07:37Z","abstract_excerpt":"Existing analyses of the expressive capacity of Transformer models have required excessively deep layers for data memorization, leading to a discrepancy with the Transformers actually used in practice. This is primarily due to the interpretation of the softmax function as an approximation of the hardmax function. By clarifying the connection between the softmax function and the Boltzmann operator, we prove that a single layer of self-attention with low-rank weight matrices possesses the capability to perfectly capture the context of an entire input sequence. As a consequence, we show that one-"},"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":"2307.14023","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-07-26T08:07:37Z","cross_cats_sorted":[],"title_canon_sha256":"e51a22fb3ed8770d8f684699f606cdcca3707da6972148c769f2b8be0e2b56f1","abstract_canon_sha256":"2ec6f2b16f8b32d8895ca2687325ede7bb4fbb5d92acafbfea44ad8ab6fea1f2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:38:29.205271Z","signature_b64":"Kj1V9eaM0rsyG1XHZ0odJnB3Z+HYhW24BZ14E0PixiauWYnYA7plYzhyLQt1jSGgeXVoUo1I/jewMgeQ5X7jCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"25e7f3d59dc55aa3eba131b7e0e100bb3bd3ba16c6657bd1216915d8b6b5ba31","last_reissued_at":"2026-07-05T07:38:29.204728Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:38:29.204728Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Are Transformers with One Layer Self-Attention Using Low-Rank Weight Matrices Universal Approximators?","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Issei Sato, Tokio Kajitsuka","submitted_at":"2023-07-26T08:07:37Z","abstract_excerpt":"Existing analyses of the expressive capacity of Transformer models have required excessively deep layers for data memorization, leading to a discrepancy with the Transformers actually used in practice. This is primarily due to the interpretation of the softmax function as an approximation of the hardmax function. By clarifying the connection between the softmax function and the Boltzmann operator, we prove that a single layer of self-attention with low-rank weight matrices possesses the capability to perfectly capture the context of an entire input sequence. As a consequence, we show that one-"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2307.14023","kind":"arxiv","version":3},"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/2307.14023/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":"2307.14023","created_at":"2026-07-05T07:38:29.204790+00:00"},{"alias_kind":"arxiv_version","alias_value":"2307.14023v3","created_at":"2026-07-05T07:38:29.204790+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2307.14023","created_at":"2026-07-05T07:38:29.204790+00:00"},{"alias_kind":"pith_short_12","alias_value":"EXT7HVM5YVNK","created_at":"2026-07-05T07:38:29.204790+00:00"},{"alias_kind":"pith_short_16","alias_value":"EXT7HVM5YVNKH25B","created_at":"2026-07-05T07:38:29.204790+00:00"},{"alias_kind":"pith_short_8","alias_value":"EXT7HVM5","created_at":"2026-07-05T07:38:29.204790+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.23364","citing_title":"Convergence of Gradient Descent for General Neural Network Architectures Beyond the NTK Regime","ref_index":111,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24878","citing_title":"Transformer Approximations from ReLUs","ref_index":3,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EXT7HVM5YVNKH25BGG36BYIAXM","json":"https://pith.science/pith/EXT7HVM5YVNKH25BGG36BYIAXM.json","graph_json":"https://pith.science/api/pith-number/EXT7HVM5YVNKH25BGG36BYIAXM/graph.json","events_json":"https://pith.science/api/pith-number/EXT7HVM5YVNKH25BGG36BYIAXM/events.json","paper":"https://pith.science/paper/EXT7HVM5"},"agent_actions":{"view_html":"https://pith.science/pith/EXT7HVM5YVNKH25BGG36BYIAXM","download_json":"https://pith.science/pith/EXT7HVM5YVNKH25BGG36BYIAXM.json","view_paper":"https://pith.science/paper/EXT7HVM5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2307.14023&json=true","fetch_graph":"https://pith.science/api/pith-number/EXT7HVM5YVNKH25BGG36BYIAXM/graph.json","fetch_events":"https://pith.science/api/pith-number/EXT7HVM5YVNKH25BGG36BYIAXM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EXT7HVM5YVNKH25BGG36BYIAXM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EXT7HVM5YVNKH25BGG36BYIAXM/action/storage_attestation","attest_author":"https://pith.science/pith/EXT7HVM5YVNKH25BGG36BYIAXM/action/author_attestation","sign_citation":"https://pith.science/pith/EXT7HVM5YVNKH25BGG36BYIAXM/action/citation_signature","submit_replication":"https://pith.science/pith/EXT7HVM5YVNKH25BGG36BYIAXM/action/replication_record"}},"created_at":"2026-07-05T07:38:29.204790+00:00","updated_at":"2026-07-05T07:38:29.204790+00:00"}