{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:V24JFKKCWYPBUAGCSAEORO5KXC","short_pith_number":"pith:V24JFKKC","schema_version":"1.0","canonical_sha256":"aeb892a942b61e1a00c29008e8bbaab88114f1f82cf38b76a34978bc07492012","source":{"kind":"arxiv","id":"2212.04458","version":2},"attestation_state":"computed","paper":{"title":"General-Purpose In-Context Learning by Meta-Learning Transformers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.NE","stat.ML"],"primary_cat":"cs.LG","authors_text":"James Harrison, Jascha Sohl-Dickstein, Louis Kirsch, Luke Metz","submitted_at":"2022-12-08T18:30:22Z","abstract_excerpt":"Modern machine learning requires system designers to specify aspects of the learning pipeline, such as losses, architectures, and optimizers. Meta-learning, or learning-to-learn, instead aims to learn those aspects, and promises to unlock greater capabilities with less manual effort. One particularly ambitious goal of meta-learning is to train general-purpose in-context learning algorithms from scratch, using only black-box models with minimal inductive bias. Such a model takes in training data, and produces test-set predictions across a wide range of problems, without any explicit definition "},"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":"2212.04458","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-12-08T18:30:22Z","cross_cats_sorted":["cs.AI","cs.NE","stat.ML"],"title_canon_sha256":"3d4d21e20a7bf0ebd4bd4da441fdee42a94cb80b368cedd961c67db2ee08967d","abstract_canon_sha256":"2f8a21e02641a64358d50574e9930e269637da5ec5e0b0f5923c7833daa3f826"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:31:26.483265Z","signature_b64":"ZuqjrMz7J+VOSbxAOPoLZovqWdC0gOAg5xhdVTo9KQ0lREWXSQa5vphLdbUlWqsR1nsVGdfpYCGO0BWkKbDjCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"aeb892a942b61e1a00c29008e8bbaab88114f1f82cf38b76a34978bc07492012","last_reissued_at":"2026-07-05T07:31:26.482811Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:31:26.482811Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"General-Purpose In-Context Learning by Meta-Learning Transformers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.NE","stat.ML"],"primary_cat":"cs.LG","authors_text":"James Harrison, Jascha Sohl-Dickstein, Louis Kirsch, Luke Metz","submitted_at":"2022-12-08T18:30:22Z","abstract_excerpt":"Modern machine learning requires system designers to specify aspects of the learning pipeline, such as losses, architectures, and optimizers. Meta-learning, or learning-to-learn, instead aims to learn those aspects, and promises to unlock greater capabilities with less manual effort. One particularly ambitious goal of meta-learning is to train general-purpose in-context learning algorithms from scratch, using only black-box models with minimal inductive bias. Such a model takes in training data, and produces test-set predictions across a wide range of problems, without any explicit definition "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2212.04458","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/2212.04458/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":"2212.04458","created_at":"2026-07-05T07:31:26.482867+00:00"},{"alias_kind":"arxiv_version","alias_value":"2212.04458v2","created_at":"2026-07-05T07:31:26.482867+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2212.04458","created_at":"2026-07-05T07:31:26.482867+00:00"},{"alias_kind":"pith_short_12","alias_value":"V24JFKKCWYPB","created_at":"2026-07-05T07:31:26.482867+00:00"},{"alias_kind":"pith_short_16","alias_value":"V24JFKKCWYPBUAGC","created_at":"2026-07-05T07:31:26.482867+00:00"},{"alias_kind":"pith_short_8","alias_value":"V24JFKKC","created_at":"2026-07-05T07:31:26.482867+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":7,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.20538","citing_title":"Multi-Task Bayesian In-Context Learning","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2606.06479","citing_title":"Pretraining Recurrent Networks without Recurrence","ref_index":68,"is_internal_anchor":false},{"citing_arxiv_id":"2605.24410","citing_title":"Advancing Graph Few-Shot Learning via In-Context Learning","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18830","citing_title":"In-Context Learning Operates as Concept Subspace Learning","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2602.01219","citing_title":"Mixture-of-Top-k Attention: Efficient Attention via Scalable Fast Weights","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2309.16797","citing_title":"Promptbreeder: Self-Referential Self-Improvement Via Prompt Evolution","ref_index":182,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05115","citing_title":"Manifold Steering Reveals the Shared Geometry of Neural Network Representation and Behavior","ref_index":292,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/V24JFKKCWYPBUAGCSAEORO5KXC","json":"https://pith.science/pith/V24JFKKCWYPBUAGCSAEORO5KXC.json","graph_json":"https://pith.science/api/pith-number/V24JFKKCWYPBUAGCSAEORO5KXC/graph.json","events_json":"https://pith.science/api/pith-number/V24JFKKCWYPBUAGCSAEORO5KXC/events.json","paper":"https://pith.science/paper/V24JFKKC"},"agent_actions":{"view_html":"https://pith.science/pith/V24JFKKCWYPBUAGCSAEORO5KXC","download_json":"https://pith.science/pith/V24JFKKCWYPBUAGCSAEORO5KXC.json","view_paper":"https://pith.science/paper/V24JFKKC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2212.04458&json=true","fetch_graph":"https://pith.science/api/pith-number/V24JFKKCWYPBUAGCSAEORO5KXC/graph.json","fetch_events":"https://pith.science/api/pith-number/V24JFKKCWYPBUAGCSAEORO5KXC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/V24JFKKCWYPBUAGCSAEORO5KXC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/V24JFKKCWYPBUAGCSAEORO5KXC/action/storage_attestation","attest_author":"https://pith.science/pith/V24JFKKCWYPBUAGCSAEORO5KXC/action/author_attestation","sign_citation":"https://pith.science/pith/V24JFKKCWYPBUAGCSAEORO5KXC/action/citation_signature","submit_replication":"https://pith.science/pith/V24JFKKCWYPBUAGCSAEORO5KXC/action/replication_record"}},"created_at":"2026-07-05T07:31:26.482867+00:00","updated_at":"2026-07-05T07:31:26.482867+00:00"}