{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:DHVMQIUX2O4TJNZMNBERTFEGA6","short_pith_number":"pith:DHVMQIUX","canonical_record":{"source":{"id":"2402.01258","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2024-02-02T09:29:40Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"54f5bc1d951657a6a5337f1e1b43356c7091c2b9af4fa4e27d86ae85aef7f250","abstract_canon_sha256":"14e617b6cb63dc8c5b5a8e3b13b64d28ea0b4befa7250e3e3cb5bf12476cff2c"},"schema_version":"1.0"},"canonical_sha256":"19eac82297d3b934b72c6849199486079a60761ad22517a27281a3dfb5c711be","source":{"kind":"arxiv","id":"2402.01258","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2402.01258","created_at":"2026-07-05T08:26:18Z"},{"alias_kind":"arxiv_version","alias_value":"2402.01258v2","created_at":"2026-07-05T08:26:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.01258","created_at":"2026-07-05T08:26:18Z"},{"alias_kind":"pith_short_12","alias_value":"DHVMQIUX2O4T","created_at":"2026-07-05T08:26:18Z"},{"alias_kind":"pith_short_16","alias_value":"DHVMQIUX2O4TJNZM","created_at":"2026-07-05T08:26:18Z"},{"alias_kind":"pith_short_8","alias_value":"DHVMQIUX","created_at":"2026-07-05T08:26:18Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:DHVMQIUX2O4TJNZMNBERTFEGA6","target":"record","payload":{"canonical_record":{"source":{"id":"2402.01258","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2024-02-02T09:29:40Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"54f5bc1d951657a6a5337f1e1b43356c7091c2b9af4fa4e27d86ae85aef7f250","abstract_canon_sha256":"14e617b6cb63dc8c5b5a8e3b13b64d28ea0b4befa7250e3e3cb5bf12476cff2c"},"schema_version":"1.0"},"canonical_sha256":"19eac82297d3b934b72c6849199486079a60761ad22517a27281a3dfb5c711be","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:26:18.518808Z","signature_b64":"Gcb5oa5+46V331uUKRG2j+rRoQpxITYZju64xBZ2Lt/jhFm3bo6JyqqK7vlCs4H3G3MtipDEpv9WY4tr7p0HDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"19eac82297d3b934b72c6849199486079a60761ad22517a27281a3dfb5c711be","last_reissued_at":"2026-07-05T08:26:18.518348Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:26:18.518348Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2402.01258","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:26:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"VFF5LIoArxvP++kE9/OkSWkrge0SbWUFtnJw23quy1GjUEQWAtNNk37g5Turnr2rpgQGlA0xfemMKuOggdd7Cg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T19:58:43.012426Z"},"content_sha256":"f96af1625bc59d4f42a1164150fb07a773267bf65685c5278b811a02c5d8725d","schema_version":"1.0","event_id":"sha256:f96af1625bc59d4f42a1164150fb07a773267bf65685c5278b811a02c5d8725d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:DHVMQIUX2O4TJNZMNBERTFEGA6","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Transformers Learn Nonlinear Features In Context: Nonconvex Mean-field Dynamics on the Attention Landscape","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"stat.ML","authors_text":"Juno Kim, Taiji Suzuki","submitted_at":"2024-02-02T09:29:40Z","abstract_excerpt":"Large language models based on the Transformer architecture have demonstrated impressive capabilities to learn in context. However, existing theoretical studies on how this phenomenon arises are limited to the dynamics of a single layer of attention trained on linear regression tasks. In this paper, we study the optimization of a Transformer consisting of a fully connected layer followed by a linear attention layer. The MLP acts as a common nonlinear representation or feature map, greatly enhancing the power of in-context learning. We prove in the mean-field and two-timescale limit that the in"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.01258","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/2402.01258/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:26:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5Ik7z2nThQfMDe8EhULLeprF/WOgcX0sYRmRxv70Ejcc/UlXHN5GcAf4a7XXqWlMUKlVEBtJLnBcrPXu6hQPBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T19:58:43.013243Z"},"content_sha256":"30fc244587bc66179624b2913b5de6cf492df73611a57df9a7f3ce47b1dfea66","schema_version":"1.0","event_id":"sha256:30fc244587bc66179624b2913b5de6cf492df73611a57df9a7f3ce47b1dfea66"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/DHVMQIUX2O4TJNZMNBERTFEGA6/bundle.json","state_url":"https://pith.science/pith/DHVMQIUX2O4TJNZMNBERTFEGA6/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/DHVMQIUX2O4TJNZMNBERTFEGA6/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-04T19:58:43Z","links":{"resolver":"https://pith.science/pith/DHVMQIUX2O4TJNZMNBERTFEGA6","bundle":"https://pith.science/pith/DHVMQIUX2O4TJNZMNBERTFEGA6/bundle.json","state":"https://pith.science/pith/DHVMQIUX2O4TJNZMNBERTFEGA6/state.json","well_known_bundle":"https://pith.science/.well-known/pith/DHVMQIUX2O4TJNZMNBERTFEGA6/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:DHVMQIUX2O4TJNZMNBERTFEGA6","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":"14e617b6cb63dc8c5b5a8e3b13b64d28ea0b4befa7250e3e3cb5bf12476cff2c","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2024-02-02T09:29:40Z","title_canon_sha256":"54f5bc1d951657a6a5337f1e1b43356c7091c2b9af4fa4e27d86ae85aef7f250"},"schema_version":"1.0","source":{"id":"2402.01258","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2402.01258","created_at":"2026-07-05T08:26:18Z"},{"alias_kind":"arxiv_version","alias_value":"2402.01258v2","created_at":"2026-07-05T08:26:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.01258","created_at":"2026-07-05T08:26:18Z"},{"alias_kind":"pith_short_12","alias_value":"DHVMQIUX2O4T","created_at":"2026-07-05T08:26:18Z"},{"alias_kind":"pith_short_16","alias_value":"DHVMQIUX2O4TJNZM","created_at":"2026-07-05T08:26:18Z"},{"alias_kind":"pith_short_8","alias_value":"DHVMQIUX","created_at":"2026-07-05T08:26:18Z"}],"graph_snapshots":[{"event_id":"sha256:30fc244587bc66179624b2913b5de6cf492df73611a57df9a7f3ce47b1dfea66","target":"graph","created_at":"2026-07-05T08:26:18Z","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/2402.01258/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large language models based on the Transformer architecture have demonstrated impressive capabilities to learn in context. However, existing theoretical studies on how this phenomenon arises are limited to the dynamics of a single layer of attention trained on linear regression tasks. In this paper, we study the optimization of a Transformer consisting of a fully connected layer followed by a linear attention layer. The MLP acts as a common nonlinear representation or feature map, greatly enhancing the power of in-context learning. We prove in the mean-field and two-timescale limit that the in","authors_text":"Juno Kim, Taiji Suzuki","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2024-02-02T09:29:40Z","title":"Transformers Learn Nonlinear Features In Context: Nonconvex Mean-field Dynamics on the Attention Landscape"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.01258","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:f96af1625bc59d4f42a1164150fb07a773267bf65685c5278b811a02c5d8725d","target":"record","created_at":"2026-07-05T08:26:18Z","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":"14e617b6cb63dc8c5b5a8e3b13b64d28ea0b4befa7250e3e3cb5bf12476cff2c","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"stat.ML","submitted_at":"2024-02-02T09:29:40Z","title_canon_sha256":"54f5bc1d951657a6a5337f1e1b43356c7091c2b9af4fa4e27d86ae85aef7f250"},"schema_version":"1.0","source":{"id":"2402.01258","kind":"arxiv","version":2}},"canonical_sha256":"19eac82297d3b934b72c6849199486079a60761ad22517a27281a3dfb5c711be","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"19eac82297d3b934b72c6849199486079a60761ad22517a27281a3dfb5c711be","first_computed_at":"2026-07-05T08:26:18.518348Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:26:18.518348Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Gcb5oa5+46V331uUKRG2j+rRoQpxITYZju64xBZ2Lt/jhFm3bo6JyqqK7vlCs4H3G3MtipDEpv9WY4tr7p0HDA==","signature_status":"signed_v1","signed_at":"2026-07-05T08:26:18.518808Z","signed_message":"canonical_sha256_bytes"},"source_id":"2402.01258","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f96af1625bc59d4f42a1164150fb07a773267bf65685c5278b811a02c5d8725d","sha256:30fc244587bc66179624b2913b5de6cf492df73611a57df9a7f3ce47b1dfea66"],"state_sha256":"9f1ccdce72c4808f7cec23cfe00971246593cc4864532bc76055c4e612218110"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"wnZjeuZBFAEBPVoZHqgl+sU30s+I5QNOxlhjNEY2SOu/bdnlEJ0YElrSb1ZU8mv4ER6fnjBU/602l1D7oKMQDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T19:58:43.019634Z","bundle_sha256":"b4624d5a59e7b8586aade3924940ca5ab19d18b58d62167b52fd22d2e416d91f"}}