{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:YZXYXTUN6G3TU2ORTZOUM2X646","short_pith_number":"pith:YZXYXTUN","schema_version":"1.0","canonical_sha256":"c66f8bce8df1b73a69d19e5d466afee7b8bc40f85a1b3ee16d47949e168f4708","source":{"kind":"arxiv","id":"2411.11465","version":4},"attestation_state":"computed","paper":{"title":"Re-examining learning linear functions in context","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Guilhem Fouilh\\'e, Nicholas Asher, Omar Naim","submitted_at":"2024-11-18T10:58:46Z","abstract_excerpt":"In-context learning (ICL) has emerged as a powerful paradigm for easily adapting Large Language Models (LLMs) to various tasks. However, our understanding of how ICL works remains limited. We explore a simple model of ICL in a controlled setup with synthetic training data to investigate ICL of univariate linear functions. We experiment with a range of GPT-2-like transformer models trained from scratch. Our findings challenge the prevailing narrative that transformers adopt algorithmic approaches like linear regression to learn a linear function in-context. These models fail to generalize beyon"},"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":"2411.11465","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-18T10:58:46Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"de37abac13b0b4327ca9fdcacf86d8ac44adf440d15f0874b221ca742955d899","abstract_canon_sha256":"5908babe4736cdde46d591b4ef9ef829f811181011ad75d90463f49bf24a7ba1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:01:55.785844Z","signature_b64":"46ThnJZWPMBs3gjsjNqxygR6mBYaKa/qscU3qdVzesDlcMIHIV3ddHSqILUQ8V8BN6mmTCfqwENa/KMUKLlzAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c66f8bce8df1b73a69d19e5d466afee7b8bc40f85a1b3ee16d47949e168f4708","last_reissued_at":"2026-07-05T12:01:55.785327Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:01:55.785327Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Re-examining learning linear functions in context","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Guilhem Fouilh\\'e, Nicholas Asher, Omar Naim","submitted_at":"2024-11-18T10:58:46Z","abstract_excerpt":"In-context learning (ICL) has emerged as a powerful paradigm for easily adapting Large Language Models (LLMs) to various tasks. However, our understanding of how ICL works remains limited. We explore a simple model of ICL in a controlled setup with synthetic training data to investigate ICL of univariate linear functions. We experiment with a range of GPT-2-like transformer models trained from scratch. Our findings challenge the prevailing narrative that transformers adopt algorithmic approaches like linear regression to learn a linear function in-context. These models fail to generalize beyon"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.11465","kind":"arxiv","version":4},"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/2411.11465/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":"2411.11465","created_at":"2026-07-05T12:01:55.785389+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.11465v4","created_at":"2026-07-05T12:01:55.785389+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.11465","created_at":"2026-07-05T12:01:55.785389+00:00"},{"alias_kind":"pith_short_12","alias_value":"YZXYXTUN6G3T","created_at":"2026-07-05T12:01:55.785389+00:00"},{"alias_kind":"pith_short_16","alias_value":"YZXYXTUN6G3TU2OR","created_at":"2026-07-05T12:01:55.785389+00:00"},{"alias_kind":"pith_short_8","alias_value":"YZXYXTUN","created_at":"2026-07-05T12:01:55.785389+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2508.14685","citing_title":"SSA: Improving Performance With a Better Scoring Function","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2509.04470","citing_title":"COCORELI: Enforcing Execution Preconditions for Reliable Collaborative Instruction Following","ref_index":2,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YZXYXTUN6G3TU2ORTZOUM2X646","json":"https://pith.science/pith/YZXYXTUN6G3TU2ORTZOUM2X646.json","graph_json":"https://pith.science/api/pith-number/YZXYXTUN6G3TU2ORTZOUM2X646/graph.json","events_json":"https://pith.science/api/pith-number/YZXYXTUN6G3TU2ORTZOUM2X646/events.json","paper":"https://pith.science/paper/YZXYXTUN"},"agent_actions":{"view_html":"https://pith.science/pith/YZXYXTUN6G3TU2ORTZOUM2X646","download_json":"https://pith.science/pith/YZXYXTUN6G3TU2ORTZOUM2X646.json","view_paper":"https://pith.science/paper/YZXYXTUN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.11465&json=true","fetch_graph":"https://pith.science/api/pith-number/YZXYXTUN6G3TU2ORTZOUM2X646/graph.json","fetch_events":"https://pith.science/api/pith-number/YZXYXTUN6G3TU2ORTZOUM2X646/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YZXYXTUN6G3TU2ORTZOUM2X646/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YZXYXTUN6G3TU2ORTZOUM2X646/action/storage_attestation","attest_author":"https://pith.science/pith/YZXYXTUN6G3TU2ORTZOUM2X646/action/author_attestation","sign_citation":"https://pith.science/pith/YZXYXTUN6G3TU2ORTZOUM2X646/action/citation_signature","submit_replication":"https://pith.science/pith/YZXYXTUN6G3TU2ORTZOUM2X646/action/replication_record"}},"created_at":"2026-07-05T12:01:55.785389+00:00","updated_at":"2026-07-05T12:01:55.785389+00:00"}