{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:K5OFQJOGJ3THYIVK7WECBJ3N3Y","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"40298e9cdd4c2a851dc2088496432dc6447beb057570e942107c454c17c98b34","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-09-11T22:42:50Z","title_canon_sha256":"84dd6d07bd43ef66e1a575e3a438b9df70306161d8e498c6826d881100a0263f"},"schema_version":"1.0","source":{"id":"2309.05858","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2309.05858","created_at":"2026-07-05T09:20:37Z"},{"alias_kind":"arxiv_version","alias_value":"2309.05858v2","created_at":"2026-07-05T09:20:37Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.05858","created_at":"2026-07-05T09:20:37Z"},{"alias_kind":"pith_short_12","alias_value":"K5OFQJOGJ3TH","created_at":"2026-07-05T09:20:37Z"},{"alias_kind":"pith_short_16","alias_value":"K5OFQJOGJ3THYIVK","created_at":"2026-07-05T09:20:37Z"},{"alias_kind":"pith_short_8","alias_value":"K5OFQJOG","created_at":"2026-07-05T09:20:37Z"}],"graph_snapshots":[{"event_id":"sha256:e13ad432d3bf67a22c4509e9192481c8dc04d7744dd21dab60cfef26451431ee","target":"graph","created_at":"2026-07-05T09:20:37Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2309.05858/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Some autoregressive models exhibit in-context learning capabilities: being able to learn as an input sequence is processed, without undergoing any parameter changes, and without being explicitly trained to do so. The origins of this phenomenon are still poorly understood. Here we analyze a series of Transformer models trained to perform synthetic sequence prediction tasks, and discover that standard next-token prediction error minimization gives rise to a subsidiary learning algorithm that adjusts the model as new inputs are revealed. We show that this process corresponds to gradient-based opt","authors_text":"Alexander Meulemans, Blaise Ag\\\"uera y Arcas, Eyvind Niklasson, Jo\\~ao Sacramento, Johannes von Oswald, Mark Sandler, Maximilian Schlegel, Max Vladymyrov, Nicolas Zucchet, Nino Scherrer, Nolan Miller, Razvan Pascanu, Seijin Kobayashi","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-09-11T22:42:50Z","title":"Uncovering mesa-optimization algorithms in Transformers"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.05858","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:438df8d9a050a6b4eaaca29c3514292a22484887cc512adfb8e32f2c460683b2","target":"record","created_at":"2026-07-05T09:20:37Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"40298e9cdd4c2a851dc2088496432dc6447beb057570e942107c454c17c98b34","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-09-11T22:42:50Z","title_canon_sha256":"84dd6d07bd43ef66e1a575e3a438b9df70306161d8e498c6826d881100a0263f"},"schema_version":"1.0","source":{"id":"2309.05858","kind":"arxiv","version":2}},"canonical_sha256":"575c5825c64ee67c22aafd8820a76dde2aba01c0fedb7b791438833e42f12463","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"575c5825c64ee67c22aafd8820a76dde2aba01c0fedb7b791438833e42f12463","first_computed_at":"2026-07-05T09:20:37.569830Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:20:37.569830Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"W658WDWiQLCVR7OFoVQ4mfjopiP+Hqp+uuvGmmLDHs0BBVm0r3me6r4UUV2gxRAjwrsHLoo8R0CqGiF7/DgwBw==","signature_status":"signed_v1","signed_at":"2026-07-05T09:20:37.570233Z","signed_message":"canonical_sha256_bytes"},"source_id":"2309.05858","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:438df8d9a050a6b4eaaca29c3514292a22484887cc512adfb8e32f2c460683b2","sha256:e13ad432d3bf67a22c4509e9192481c8dc04d7744dd21dab60cfef26451431ee"],"state_sha256":"8c6d3172e0f4f5ed0736e4dc7e9b34c54522117f3748e7970fa4a7c9acbd7403"}