{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2022:XJS4KDVO5PSVP3L76RXU7RDLGE","short_pith_number":"pith:XJS4KDVO","canonical_record":{"source":{"id":"2212.10559","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-12-20T18:58:48Z","cross_cats_sorted":[],"title_canon_sha256":"85d5aae217fc86bf6ff9e7bd1e1a30ae1be7a7d461df81fcf885830126b163a9","abstract_canon_sha256":"5572a489d43f509b6ed9fc86fb460edff1b60e1cb01b66aebe0669af7d091d60"},"schema_version":"1.0"},"canonical_sha256":"ba65c50eaeebe557ed7ff46f4fc46b31112645d3e7be804f6f64309742d4a55d","source":{"kind":"arxiv","id":"2212.10559","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2212.10559","created_at":"2026-07-05T06:09:47Z"},{"alias_kind":"arxiv_version","alias_value":"2212.10559v3","created_at":"2026-07-05T06:09:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2212.10559","created_at":"2026-07-05T06:09:47Z"},{"alias_kind":"pith_short_12","alias_value":"XJS4KDVO5PSV","created_at":"2026-07-05T06:09:47Z"},{"alias_kind":"pith_short_16","alias_value":"XJS4KDVO5PSVP3L7","created_at":"2026-07-05T06:09:47Z"},{"alias_kind":"pith_short_8","alias_value":"XJS4KDVO","created_at":"2026-07-05T06:09:47Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2022:XJS4KDVO5PSVP3L76RXU7RDLGE","target":"record","payload":{"canonical_record":{"source":{"id":"2212.10559","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-12-20T18:58:48Z","cross_cats_sorted":[],"title_canon_sha256":"85d5aae217fc86bf6ff9e7bd1e1a30ae1be7a7d461df81fcf885830126b163a9","abstract_canon_sha256":"5572a489d43f509b6ed9fc86fb460edff1b60e1cb01b66aebe0669af7d091d60"},"schema_version":"1.0"},"canonical_sha256":"ba65c50eaeebe557ed7ff46f4fc46b31112645d3e7be804f6f64309742d4a55d","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:09:47.320219Z","signature_b64":"F1QXFAa574HBFCWlS61v7td0ZVwe5/uiQ8D+S0HpEQg+fs8MnVg0L6JGZ2+AWZq4LB0s13LhTBEwbskzRgz4BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ba65c50eaeebe557ed7ff46f4fc46b31112645d3e7be804f6f64309742d4a55d","last_reissued_at":"2026-07-05T06:09:47.319664Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:09:47.319664Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2212.10559","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T06:09:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"epJUcza80sxzCR9fYOVQCOsnajG4Oi8nCgI3rwTKpbEHtkjUvxIOn9v9QrRFCW43dmC/QOUOrAuQbZCX7CT2DQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-25T05:23:46.888572Z"},"content_sha256":"fe1cff5860a427929b20407b2f41d2475e4629ef64e95a381aad4b95c738a130","schema_version":"1.0","event_id":"sha256:fe1cff5860a427929b20407b2f41d2475e4629ef64e95a381aad4b95c738a130"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2022:XJS4KDVO5PSVP3L76RXU7RDLGE","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Why Can GPT Learn In-Context? Language Models Implicitly Perform Gradient Descent as Meta-Optimizers","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Damai Dai, Furu Wei, Li Dong, Shuming Ma, Yaru Hao, Yutao Sun, Zhifang Sui","submitted_at":"2022-12-20T18:58:48Z","abstract_excerpt":"Large pretrained language models have shown surprising in-context learning (ICL) ability. With a few demonstration input-label pairs, they can predict the label for an unseen input without parameter updates. Despite the great success in performance, its working mechanism still remains an open question. In this paper, we explain language models as meta-optimizers and understand in-context learning as implicit finetuning. Theoretically, we figure out that Transformer attention has a dual form of gradient descent. On top of it, we understand ICL as follows: GPT first produces meta-gradients accor"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2212.10559","kind":"arxiv","version":3},"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/2212.10559/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":"2026-07-05T06:09:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"yqFp18nhO9+vH6XdOuNI6YbMgEGLUuQ1Rtv07K+ukmpiuqSFsIhGu95qtw39nGv+dW1Ow+Sw5Y5njuneYHinBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-25T05:23:46.888954Z"},"content_sha256":"3653f9a490157117192f2ff817257f2b2ca6b50c43a935c7875554c2a51f7237","schema_version":"1.0","event_id":"sha256:3653f9a490157117192f2ff817257f2b2ca6b50c43a935c7875554c2a51f7237"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/XJS4KDVO5PSVP3L76RXU7RDLGE/bundle.json","state_url":"https://pith.science/pith/XJS4KDVO5PSVP3L76RXU7RDLGE/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/XJS4KDVO5PSVP3L76RXU7RDLGE/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":"2026-07-25T05:23:46Z","links":{"resolver":"https://pith.science/pith/XJS4KDVO5PSVP3L76RXU7RDLGE","bundle":"https://pith.science/pith/XJS4KDVO5PSVP3L76RXU7RDLGE/bundle.json","state":"https://pith.science/pith/XJS4KDVO5PSVP3L76RXU7RDLGE/state.json","well_known_bundle":"https://pith.science/.well-known/pith/XJS4KDVO5PSVP3L76RXU7RDLGE/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:XJS4KDVO5PSVP3L76RXU7RDLGE","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":"5572a489d43f509b6ed9fc86fb460edff1b60e1cb01b66aebe0669af7d091d60","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-12-20T18:58:48Z","title_canon_sha256":"85d5aae217fc86bf6ff9e7bd1e1a30ae1be7a7d461df81fcf885830126b163a9"},"schema_version":"1.0","source":{"id":"2212.10559","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2212.10559","created_at":"2026-07-05T06:09:47Z"},{"alias_kind":"arxiv_version","alias_value":"2212.10559v3","created_at":"2026-07-05T06:09:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2212.10559","created_at":"2026-07-05T06:09:47Z"},{"alias_kind":"pith_short_12","alias_value":"XJS4KDVO5PSV","created_at":"2026-07-05T06:09:47Z"},{"alias_kind":"pith_short_16","alias_value":"XJS4KDVO5PSVP3L7","created_at":"2026-07-05T06:09:47Z"},{"alias_kind":"pith_short_8","alias_value":"XJS4KDVO","created_at":"2026-07-05T06:09:47Z"}],"graph_snapshots":[{"event_id":"sha256:3653f9a490157117192f2ff817257f2b2ca6b50c43a935c7875554c2a51f7237","target":"graph","created_at":"2026-07-05T06:09:47Z","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/2212.10559/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large pretrained language models have shown surprising in-context learning (ICL) ability. With a few demonstration input-label pairs, they can predict the label for an unseen input without parameter updates. Despite the great success in performance, its working mechanism still remains an open question. In this paper, we explain language models as meta-optimizers and understand in-context learning as implicit finetuning. Theoretically, we figure out that Transformer attention has a dual form of gradient descent. On top of it, we understand ICL as follows: GPT first produces meta-gradients accor","authors_text":"Damai Dai, Furu Wei, Li Dong, Shuming Ma, Yaru Hao, Yutao Sun, Zhifang Sui","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-12-20T18:58:48Z","title":"Why Can GPT Learn In-Context? Language Models Implicitly Perform Gradient Descent as Meta-Optimizers"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2212.10559","kind":"arxiv","version":3},"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:fe1cff5860a427929b20407b2f41d2475e4629ef64e95a381aad4b95c738a130","target":"record","created_at":"2026-07-05T06:09:47Z","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":"5572a489d43f509b6ed9fc86fb460edff1b60e1cb01b66aebe0669af7d091d60","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2022-12-20T18:58:48Z","title_canon_sha256":"85d5aae217fc86bf6ff9e7bd1e1a30ae1be7a7d461df81fcf885830126b163a9"},"schema_version":"1.0","source":{"id":"2212.10559","kind":"arxiv","version":3}},"canonical_sha256":"ba65c50eaeebe557ed7ff46f4fc46b31112645d3e7be804f6f64309742d4a55d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ba65c50eaeebe557ed7ff46f4fc46b31112645d3e7be804f6f64309742d4a55d","first_computed_at":"2026-07-05T06:09:47.319664Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T06:09:47.319664Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"F1QXFAa574HBFCWlS61v7td0ZVwe5/uiQ8D+S0HpEQg+fs8MnVg0L6JGZ2+AWZq4LB0s13LhTBEwbskzRgz4BA==","signature_status":"signed_v1","signed_at":"2026-07-05T06:09:47.320219Z","signed_message":"canonical_sha256_bytes"},"source_id":"2212.10559","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:fe1cff5860a427929b20407b2f41d2475e4629ef64e95a381aad4b95c738a130","sha256:3653f9a490157117192f2ff817257f2b2ca6b50c43a935c7875554c2a51f7237"],"state_sha256":"94e7969ecca61716e7013c2009d5aed00e949f3a591f0104adcaf8e274ecebd0"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"XebnA4BC1P8HRvVq2kjvjvru4VOm0jhTP3vdXc1MAe07w8IeShil6EhUxC29CtwDMU6jWwch1TnVDzebsrtfCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-25T05:23:46.891746Z","bundle_sha256":"2c7494ead6433fc4125f5ef48040e037d9fc19388728aa6a2a429169f6749476"}}