{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:FFCSP4PZ2IWXSIZWG4NVBJUCLX","short_pith_number":"pith:FFCSP4PZ","schema_version":"1.0","canonical_sha256":"294527f1f9d22d792336371b50a6825dcefa833d1d01637c31a41f627c161df4","source":{"kind":"arxiv","id":"2408.12186","version":2},"attestation_state":"computed","paper":{"title":"Transformers are Minimax Optimal Nonparametric In-Context Learners","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Juno Kim, Taiji Suzuki, Tai Nakamaki","submitted_at":"2024-08-22T08:02:10Z","abstract_excerpt":"In-context learning (ICL) of large language models has proven to be a surprisingly effective method of learning a new task from only a few demonstrative examples. In this paper, we study the efficacy of ICL from the viewpoint of statistical learning theory. We develop approximation and generalization error bounds for a transformer composed of a deep neural network and one linear attention layer, pretrained on nonparametric regression tasks sampled from general function spaces including the Besov space and piecewise $\\gamma$-smooth class. We show that sufficiently trained transformers can achie"},"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":"2408.12186","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2024-08-22T08:02:10Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"6b245a21ca86f4e7e111a0fb9b4654fb1b5b4493198b770c3c97fd055fb12020","abstract_canon_sha256":"2903b9d8b9608ff67c1cedb7bff88a1674aa374e6c0a16c44c11cd2721694088"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:14:37.524058Z","signature_b64":"fDJ/DwWYpQfxx/RuYKdlUI2zJs6UvNjcGLIUeT+eWsGwBSDXnRBkQR+LSSpcnblWFiecOUpj5kYox9tnqD/zBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"294527f1f9d22d792336371b50a6825dcefa833d1d01637c31a41f627c161df4","last_reissued_at":"2026-07-05T09:14:37.523533Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:14:37.523533Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Transformers are Minimax Optimal Nonparametric In-Context Learners","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Juno Kim, Taiji Suzuki, Tai Nakamaki","submitted_at":"2024-08-22T08:02:10Z","abstract_excerpt":"In-context learning (ICL) of large language models has proven to be a surprisingly effective method of learning a new task from only a few demonstrative examples. In this paper, we study the efficacy of ICL from the viewpoint of statistical learning theory. We develop approximation and generalization error bounds for a transformer composed of a deep neural network and one linear attention layer, pretrained on nonparametric regression tasks sampled from general function spaces including the Besov space and piecewise $\\gamma$-smooth class. We show that sufficiently trained transformers can achie"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.12186","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/2408.12186/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":"2408.12186","created_at":"2026-07-05T09:14:37.523606+00:00"},{"alias_kind":"arxiv_version","alias_value":"2408.12186v2","created_at":"2026-07-05T09:14:37.523606+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.12186","created_at":"2026-07-05T09:14:37.523606+00:00"},{"alias_kind":"pith_short_12","alias_value":"FFCSP4PZ2IWX","created_at":"2026-07-05T09:14:37.523606+00:00"},{"alias_kind":"pith_short_16","alias_value":"FFCSP4PZ2IWXSIZW","created_at":"2026-07-05T09:14:37.523606+00:00"},{"alias_kind":"pith_short_8","alias_value":"FFCSP4PZ","created_at":"2026-07-05T09:14:37.523606+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2502.12187","citing_title":"Hallucinations are inevitable but can be made statistically negligible","ref_index":55,"is_internal_anchor":false},{"citing_arxiv_id":"2506.00182","citing_title":"Overfitting has a limitation: a model-independent generalization gap bound based on R\\'enyi entropy","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20271","citing_title":"Multi-Head Attention as Ensemble Nadaraya-Watson Estimation: Variance Reduction, Decorrelation, and Optimal Head Diversity","ref_index":11,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FFCSP4PZ2IWXSIZWG4NVBJUCLX","json":"https://pith.science/pith/FFCSP4PZ2IWXSIZWG4NVBJUCLX.json","graph_json":"https://pith.science/api/pith-number/FFCSP4PZ2IWXSIZWG4NVBJUCLX/graph.json","events_json":"https://pith.science/api/pith-number/FFCSP4PZ2IWXSIZWG4NVBJUCLX/events.json","paper":"https://pith.science/paper/FFCSP4PZ"},"agent_actions":{"view_html":"https://pith.science/pith/FFCSP4PZ2IWXSIZWG4NVBJUCLX","download_json":"https://pith.science/pith/FFCSP4PZ2IWXSIZWG4NVBJUCLX.json","view_paper":"https://pith.science/paper/FFCSP4PZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2408.12186&json=true","fetch_graph":"https://pith.science/api/pith-number/FFCSP4PZ2IWXSIZWG4NVBJUCLX/graph.json","fetch_events":"https://pith.science/api/pith-number/FFCSP4PZ2IWXSIZWG4NVBJUCLX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FFCSP4PZ2IWXSIZWG4NVBJUCLX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FFCSP4PZ2IWXSIZWG4NVBJUCLX/action/storage_attestation","attest_author":"https://pith.science/pith/FFCSP4PZ2IWXSIZWG4NVBJUCLX/action/author_attestation","sign_citation":"https://pith.science/pith/FFCSP4PZ2IWXSIZWG4NVBJUCLX/action/citation_signature","submit_replication":"https://pith.science/pith/FFCSP4PZ2IWXSIZWG4NVBJUCLX/action/replication_record"}},"created_at":"2026-07-05T09:14:37.523606+00:00","updated_at":"2026-07-05T09:14:37.523606+00:00"}