{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:YUJTVOG6FLHVPR5NIFLKMNT3KZ","short_pith_number":"pith:YUJTVOG6","schema_version":"1.0","canonical_sha256":"c5133ab8de2acf57c7ad4156a6367b564465e311dab5789be10a283986cd1d67","source":{"kind":"arxiv","id":"2402.00743","version":2},"attestation_state":"computed","paper":{"title":"Theoretical Understanding of In-Context Learning in Shallow Transformers with Unstructured Data","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.CL","stat.ML"],"primary_cat":"cs.LG","authors_text":"Chenheng Xu, Guang Cheng, Namjoon Suh, Qifan Song, Xiaofeng Lin, Yue Xing","submitted_at":"2024-02-01T16:39:45Z","abstract_excerpt":"Large language models (LLMs) are powerful models that can learn concepts at the inference stage via in-context learning (ICL). While theoretical studies, e.g., \\cite{zhang2023trained}, attempt to explain the mechanism of ICL, they assume the input $x_i$ and the output $y_i$ of each demonstration example are in the same token (i.e., structured data). However, in real practice, the examples are usually text input, and all words, regardless of their logic relationship, are stored in different tokens (i.e., unstructured data \\cite{wibisono2023role}). To understand how LLMs learn from the unstructu"},"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":"2402.00743","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-02-01T16:39:45Z","cross_cats_sorted":["cs.CL","stat.ML"],"title_canon_sha256":"dcb6d90ace8bd284b1c72142bc1ed44b84bfbee9131df5b470da31399b1e58cc","abstract_canon_sha256":"7bd2d04decce8abae5002fa7dd07bfc104481e5a5394a51312a4faf9df94c7fa"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:33:35.108159Z","signature_b64":"XEwl2sEmN09lKwhB7fwqb1LnZ7A8pmfvnJSr/cRkOXCpgtRPRc884LjkU0PuqYE7pQ7lzU+s1Sxmer9saCd+Cg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c5133ab8de2acf57c7ad4156a6367b564465e311dab5789be10a283986cd1d67","last_reissued_at":"2026-07-05T08:33:35.107666Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:33:35.107666Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Theoretical Understanding of In-Context Learning in Shallow Transformers with Unstructured Data","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.CL","stat.ML"],"primary_cat":"cs.LG","authors_text":"Chenheng Xu, Guang Cheng, Namjoon Suh, Qifan Song, Xiaofeng Lin, Yue Xing","submitted_at":"2024-02-01T16:39:45Z","abstract_excerpt":"Large language models (LLMs) are powerful models that can learn concepts at the inference stage via in-context learning (ICL). While theoretical studies, e.g., \\cite{zhang2023trained}, attempt to explain the mechanism of ICL, they assume the input $x_i$ and the output $y_i$ of each demonstration example are in the same token (i.e., structured data). However, in real practice, the examples are usually text input, and all words, regardless of their logic relationship, are stored in different tokens (i.e., unstructured data \\cite{wibisono2023role}). To understand how LLMs learn from the unstructu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.00743","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/2402.00743/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":"2402.00743","created_at":"2026-07-05T08:33:35.107730+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.00743v2","created_at":"2026-07-05T08:33:35.107730+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.00743","created_at":"2026-07-05T08:33:35.107730+00:00"},{"alias_kind":"pith_short_12","alias_value":"YUJTVOG6FLHV","created_at":"2026-07-05T08:33:35.107730+00:00"},{"alias_kind":"pith_short_16","alias_value":"YUJTVOG6FLHVPR5N","created_at":"2026-07-05T08:33:35.107730+00:00"},{"alias_kind":"pith_short_8","alias_value":"YUJTVOG6","created_at":"2026-07-05T08:33:35.107730+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.09048","citing_title":"Understanding Task Vectors in In-Context Learning: Emergence, Functionality, and Limitations","ref_index":22,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YUJTVOG6FLHVPR5NIFLKMNT3KZ","json":"https://pith.science/pith/YUJTVOG6FLHVPR5NIFLKMNT3KZ.json","graph_json":"https://pith.science/api/pith-number/YUJTVOG6FLHVPR5NIFLKMNT3KZ/graph.json","events_json":"https://pith.science/api/pith-number/YUJTVOG6FLHVPR5NIFLKMNT3KZ/events.json","paper":"https://pith.science/paper/YUJTVOG6"},"agent_actions":{"view_html":"https://pith.science/pith/YUJTVOG6FLHVPR5NIFLKMNT3KZ","download_json":"https://pith.science/pith/YUJTVOG6FLHVPR5NIFLKMNT3KZ.json","view_paper":"https://pith.science/paper/YUJTVOG6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.00743&json=true","fetch_graph":"https://pith.science/api/pith-number/YUJTVOG6FLHVPR5NIFLKMNT3KZ/graph.json","fetch_events":"https://pith.science/api/pith-number/YUJTVOG6FLHVPR5NIFLKMNT3KZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YUJTVOG6FLHVPR5NIFLKMNT3KZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YUJTVOG6FLHVPR5NIFLKMNT3KZ/action/storage_attestation","attest_author":"https://pith.science/pith/YUJTVOG6FLHVPR5NIFLKMNT3KZ/action/author_attestation","sign_citation":"https://pith.science/pith/YUJTVOG6FLHVPR5NIFLKMNT3KZ/action/citation_signature","submit_replication":"https://pith.science/pith/YUJTVOG6FLHVPR5NIFLKMNT3KZ/action/replication_record"}},"created_at":"2026-07-05T08:33:35.107730+00:00","updated_at":"2026-07-05T08:33:35.107730+00:00"}