{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:AQGRVWJIK3YBRI3DZHATNQLUAD","short_pith_number":"pith:AQGRVWJI","schema_version":"1.0","canonical_sha256":"040d1ad92856f018a363c9c136c17400de6a3e4941c0e4bc1a123e782a8f7a2a","source":{"kind":"arxiv","id":"2412.01003","version":4},"attestation_state":"computed","paper":{"title":"Competition Dynamics Shape Algorithmic Phases of In-Context Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Core Francisco Park, Ekdeep Singh Lubana, Hidenori Tanaka, Itamar Pres","submitted_at":"2024-12-01T23:35:53Z","abstract_excerpt":"In-Context Learning (ICL) has significantly expanded the general-purpose nature of large language models, allowing them to adapt to novel tasks using merely the inputted context. This has motivated a series of papers that analyze tractable synthetic domains and postulate precise mechanisms that may underlie ICL. However, the use of relatively distinct setups that often lack a sequence modeling nature to them makes it unclear how general the reported insights from such studies are. Motivated by this, we propose a synthetic sequence modeling task that involves learning to simulate a finite mixtu"},"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":"2412.01003","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-12-01T23:35:53Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"4700a6119aef279ef707196cb882c1c22a7cea8ad29cb3cfc0d08b4498599090","abstract_canon_sha256":"c71b20602b1d73f08643f8df46f0bfca7ad12cb1ae5199df2b6bd43f4a2e9c14"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:57:30.136892Z","signature_b64":"Rl2kk4SO9JbdUDIoekQHFeBAmyS6G+mZ/PDBHcWzoLD0Qawuvv2KlLStngBiycP8kteyumQTYgxhaFCcuX9BBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"040d1ad92856f018a363c9c136c17400de6a3e4941c0e4bc1a123e782a8f7a2a","last_reissued_at":"2026-07-05T10:57:30.136347Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:57:30.136347Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Competition Dynamics Shape Algorithmic Phases of In-Context Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Core Francisco Park, Ekdeep Singh Lubana, Hidenori Tanaka, Itamar Pres","submitted_at":"2024-12-01T23:35:53Z","abstract_excerpt":"In-Context Learning (ICL) has significantly expanded the general-purpose nature of large language models, allowing them to adapt to novel tasks using merely the inputted context. This has motivated a series of papers that analyze tractable synthetic domains and postulate precise mechanisms that may underlie ICL. However, the use of relatively distinct setups that often lack a sequence modeling nature to them makes it unclear how general the reported insights from such studies are. Motivated by this, we propose a synthetic sequence modeling task that involves learning to simulate a finite mixtu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.01003","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/2412.01003/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":"2412.01003","created_at":"2026-07-05T10:57:30.136410+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.01003v4","created_at":"2026-07-05T10:57:30.136410+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.01003","created_at":"2026-07-05T10:57:30.136410+00:00"},{"alias_kind":"pith_short_12","alias_value":"AQGRVWJIK3YB","created_at":"2026-07-05T10:57:30.136410+00:00"},{"alias_kind":"pith_short_16","alias_value":"AQGRVWJIK3YBRI3D","created_at":"2026-07-05T10:57:30.136410+00:00"},{"alias_kind":"pith_short_8","alias_value":"AQGRVWJI","created_at":"2026-07-05T10:57:30.136410+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.05076","citing_title":"High-Dimensional Statistics: Reflections on Progress and Open Problems","ref_index":78,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07568","citing_title":"A Systematic Study of Behavioral Cloning for Scientific Data Annotation","ref_index":78,"is_internal_anchor":false},{"citing_arxiv_id":"2511.01292","citing_title":"Optimal Attention Temperature Improves the Robustness of In-Context Learning under Distribution Shift in High Dimensions","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12412","citing_title":"Stories in Space: In-Context Learning Trajectories in Conceptual Belief Space","ref_index":116,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05076","citing_title":"High-Dimensional Statistics: Reflections on Progress and Open Problems","ref_index":78,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AQGRVWJIK3YBRI3DZHATNQLUAD","json":"https://pith.science/pith/AQGRVWJIK3YBRI3DZHATNQLUAD.json","graph_json":"https://pith.science/api/pith-number/AQGRVWJIK3YBRI3DZHATNQLUAD/graph.json","events_json":"https://pith.science/api/pith-number/AQGRVWJIK3YBRI3DZHATNQLUAD/events.json","paper":"https://pith.science/paper/AQGRVWJI"},"agent_actions":{"view_html":"https://pith.science/pith/AQGRVWJIK3YBRI3DZHATNQLUAD","download_json":"https://pith.science/pith/AQGRVWJIK3YBRI3DZHATNQLUAD.json","view_paper":"https://pith.science/paper/AQGRVWJI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.01003&json=true","fetch_graph":"https://pith.science/api/pith-number/AQGRVWJIK3YBRI3DZHATNQLUAD/graph.json","fetch_events":"https://pith.science/api/pith-number/AQGRVWJIK3YBRI3DZHATNQLUAD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AQGRVWJIK3YBRI3DZHATNQLUAD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AQGRVWJIK3YBRI3DZHATNQLUAD/action/storage_attestation","attest_author":"https://pith.science/pith/AQGRVWJIK3YBRI3DZHATNQLUAD/action/author_attestation","sign_citation":"https://pith.science/pith/AQGRVWJIK3YBRI3DZHATNQLUAD/action/citation_signature","submit_replication":"https://pith.science/pith/AQGRVWJIK3YBRI3DZHATNQLUAD/action/replication_record"}},"created_at":"2026-07-05T10:57:30.136410+00:00","updated_at":"2026-07-05T10:57:30.136410+00:00"}