{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:ZOL5FKOLYKPZ5RXQ2BYMSQHQEC","short_pith_number":"pith:ZOL5FKOL","canonical_record":{"source":{"id":"2310.06112","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-10-09T19:40:25Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"8b6e3c05225d54e8b7eaaefb927ae2963a4433e31f01e4f1ce39efc1cedfe3c8","abstract_canon_sha256":"41cd7339b10df8244fbf6e8c0599b2cd314d9819ab7fea4ff6a4a635bf612c09"},"schema_version":"1.0"},"canonical_sha256":"cb97d2a9cbc29f9ec6f0d070c940f02089ffc8ae327ef4e68e1041d07ce6f4dc","source":{"kind":"arxiv","id":"2310.06112","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.06112","created_at":"2026-07-05T07:40:57Z"},{"alias_kind":"arxiv_version","alias_value":"2310.06112v2","created_at":"2026-07-05T07:40:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.06112","created_at":"2026-07-05T07:40:57Z"},{"alias_kind":"pith_short_12","alias_value":"ZOL5FKOLYKPZ","created_at":"2026-07-05T07:40:57Z"},{"alias_kind":"pith_short_16","alias_value":"ZOL5FKOLYKPZ5RXQ","created_at":"2026-07-05T07:40:57Z"},{"alias_kind":"pith_short_8","alias_value":"ZOL5FKOL","created_at":"2026-07-05T07:40:57Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:ZOL5FKOLYKPZ5RXQ2BYMSQHQEC","target":"record","payload":{"canonical_record":{"source":{"id":"2310.06112","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-10-09T19:40:25Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"8b6e3c05225d54e8b7eaaefb927ae2963a4433e31f01e4f1ce39efc1cedfe3c8","abstract_canon_sha256":"41cd7339b10df8244fbf6e8c0599b2cd314d9819ab7fea4ff6a4a635bf612c09"},"schema_version":"1.0"},"canonical_sha256":"cb97d2a9cbc29f9ec6f0d070c940f02089ffc8ae327ef4e68e1041d07ce6f4dc","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:40:57.832186Z","signature_b64":"7CFNZ/ziG/KFamonzqUPh9ezAqeVBDVfoviI7XEvnDk4N07o6T7s2pYHWUPg5fxiWicVKJTI3No2ABNxFokkAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cb97d2a9cbc29f9ec6f0d070c940f02089ffc8ae327ef4e68e1041d07ce6f4dc","last_reissued_at":"2026-07-05T07:40:57.831741Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:40:57.831741Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2310.06112","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":"2026-07-05T07:40:57Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2Nu838DRn10q6QdoAWhA2K66S2NUBxWHe+f/CY/EnwH12ka5oyDi9bgeyFxQWa/3vaXo+AKW3PKFIVcW/EgHBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T06:24:13.958579Z"},"content_sha256":"e2ec14e452f0158fadcca222a33f688d81e6938bbcf212a37011be21cf5b99e7","schema_version":"1.0","event_id":"sha256:e2ec14e452f0158fadcca222a33f688d81e6938bbcf212a37011be21cf5b99e7"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:ZOL5FKOLYKPZ5RXQ2BYMSQHQEC","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Theoretical Analysis of Robust Overfitting for Wide DNNs: An NTK Approach","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Di Wang, Shaopeng Fu","submitted_at":"2023-10-09T19:40:25Z","abstract_excerpt":"Adversarial training (AT) is a canonical method for enhancing the robustness of deep neural networks (DNNs). However, recent studies empirically demonstrated that it suffers from robust overfitting, i.e., a long time AT can be detrimental to the robustness of DNNs. This paper presents a theoretical explanation of robust overfitting for DNNs. Specifically, we non-trivially extend the neural tangent kernel (NTK) theory to AT and prove that an adversarially trained wide DNN can be well approximated by a linearized DNN. Moreover, for squared loss, closed-form AT dynamics for the linearized DNN can"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.06112","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/2310.06112/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":"2026-07-05T07:40:57Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"l6/X4F71CM0nD465BGzziJ7XQSTgJ4JZyHTqj1kQuZNq7MbaqI0Nms/Rtz3wi0gq2lHPab4gnEvreMy8DUqdCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T06:24:13.959077Z"},"content_sha256":"2febd5c30a9c103b9f50351be26a8d0cd41085186ac0548c96354660d6859b84","schema_version":"1.0","event_id":"sha256:2febd5c30a9c103b9f50351be26a8d0cd41085186ac0548c96354660d6859b84"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ZOL5FKOLYKPZ5RXQ2BYMSQHQEC/bundle.json","state_url":"https://pith.science/pith/ZOL5FKOLYKPZ5RXQ2BYMSQHQEC/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ZOL5FKOLYKPZ5RXQ2BYMSQHQEC/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":"2026-08-08T06:24:13Z","links":{"resolver":"https://pith.science/pith/ZOL5FKOLYKPZ5RXQ2BYMSQHQEC","bundle":"https://pith.science/pith/ZOL5FKOLYKPZ5RXQ2BYMSQHQEC/bundle.json","state":"https://pith.science/pith/ZOL5FKOLYKPZ5RXQ2BYMSQHQEC/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ZOL5FKOLYKPZ5RXQ2BYMSQHQEC/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:ZOL5FKOLYKPZ5RXQ2BYMSQHQEC","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":"41cd7339b10df8244fbf6e8c0599b2cd314d9819ab7fea4ff6a4a635bf612c09","cross_cats_sorted":["stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-10-09T19:40:25Z","title_canon_sha256":"8b6e3c05225d54e8b7eaaefb927ae2963a4433e31f01e4f1ce39efc1cedfe3c8"},"schema_version":"1.0","source":{"id":"2310.06112","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.06112","created_at":"2026-07-05T07:40:57Z"},{"alias_kind":"arxiv_version","alias_value":"2310.06112v2","created_at":"2026-07-05T07:40:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.06112","created_at":"2026-07-05T07:40:57Z"},{"alias_kind":"pith_short_12","alias_value":"ZOL5FKOLYKPZ","created_at":"2026-07-05T07:40:57Z"},{"alias_kind":"pith_short_16","alias_value":"ZOL5FKOLYKPZ5RXQ","created_at":"2026-07-05T07:40:57Z"},{"alias_kind":"pith_short_8","alias_value":"ZOL5FKOL","created_at":"2026-07-05T07:40:57Z"}],"graph_snapshots":[{"event_id":"sha256:2febd5c30a9c103b9f50351be26a8d0cd41085186ac0548c96354660d6859b84","target":"graph","created_at":"2026-07-05T07:40:57Z","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/2310.06112/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Adversarial training (AT) is a canonical method for enhancing the robustness of deep neural networks (DNNs). However, recent studies empirically demonstrated that it suffers from robust overfitting, i.e., a long time AT can be detrimental to the robustness of DNNs. This paper presents a theoretical explanation of robust overfitting for DNNs. Specifically, we non-trivially extend the neural tangent kernel (NTK) theory to AT and prove that an adversarially trained wide DNN can be well approximated by a linearized DNN. Moreover, for squared loss, closed-form AT dynamics for the linearized DNN can","authors_text":"Di Wang, Shaopeng Fu","cross_cats":["stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-10-09T19:40:25Z","title":"Theoretical Analysis of Robust Overfitting for Wide DNNs: An NTK Approach"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.06112","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:e2ec14e452f0158fadcca222a33f688d81e6938bbcf212a37011be21cf5b99e7","target":"record","created_at":"2026-07-05T07:40:57Z","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":"41cd7339b10df8244fbf6e8c0599b2cd314d9819ab7fea4ff6a4a635bf612c09","cross_cats_sorted":["stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-10-09T19:40:25Z","title_canon_sha256":"8b6e3c05225d54e8b7eaaefb927ae2963a4433e31f01e4f1ce39efc1cedfe3c8"},"schema_version":"1.0","source":{"id":"2310.06112","kind":"arxiv","version":2}},"canonical_sha256":"cb97d2a9cbc29f9ec6f0d070c940f02089ffc8ae327ef4e68e1041d07ce6f4dc","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"cb97d2a9cbc29f9ec6f0d070c940f02089ffc8ae327ef4e68e1041d07ce6f4dc","first_computed_at":"2026-07-05T07:40:57.831741Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:40:57.831741Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"7CFNZ/ziG/KFamonzqUPh9ezAqeVBDVfoviI7XEvnDk4N07o6T7s2pYHWUPg5fxiWicVKJTI3No2ABNxFokkAg==","signature_status":"signed_v1","signed_at":"2026-07-05T07:40:57.832186Z","signed_message":"canonical_sha256_bytes"},"source_id":"2310.06112","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e2ec14e452f0158fadcca222a33f688d81e6938bbcf212a37011be21cf5b99e7","sha256:2febd5c30a9c103b9f50351be26a8d0cd41085186ac0548c96354660d6859b84"],"state_sha256":"a6c6c0ac2102d9f7d21729a62c9f89784feb7203267d4b9b1e3d17d142b538fa"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"nOPDYeOaiP0vS5sQX1+jWYwX3XbmCsN4bVJvxOWqDhxo5O8bt4jaS8hWZ8BzC2GhaSc3Pmpczomv16tvrApJAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T06:24:13.963931Z","bundle_sha256":"50ae3473d912f20791d6c2d455bc74fee4c9a5dba958824c06a3b9b585a6c142"}}