{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2021:3P2VQI5NZUJPVBVQSIPPXKR6U7","short_pith_number":"pith:3P2VQI5N","canonical_record":{"source":{"id":"2110.02732","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-10-06T13:27:23Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"a1e3ec8df7eeb2c0d3e45fdd5a6bf5061868a52282691624eac477e30d2fc96b","abstract_canon_sha256":"3498c8682d30c7906e54683cbbc9a735eb1f8a2084034ea4445e2327acdd2a20"},"schema_version":"1.0"},"canonical_sha256":"dbf55823adcd12fa86b0921efbaa3ea7e7e6e626442fae486fbdb01a3ca7a69a","source":{"kind":"arxiv","id":"2110.02732","version":4},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2110.02732","created_at":"2026-07-05T05:02:23Z"},{"alias_kind":"arxiv_version","alias_value":"2110.02732v4","created_at":"2026-07-05T05:02:23Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2110.02732","created_at":"2026-07-05T05:02:23Z"},{"alias_kind":"pith_short_12","alias_value":"3P2VQI5NZUJP","created_at":"2026-07-05T05:02:23Z"},{"alias_kind":"pith_short_16","alias_value":"3P2VQI5NZUJPVBVQ","created_at":"2026-07-05T05:02:23Z"},{"alias_kind":"pith_short_8","alias_value":"3P2VQI5N","created_at":"2026-07-05T05:02:23Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2021:3P2VQI5NZUJPVBVQSIPPXKR6U7","target":"record","payload":{"canonical_record":{"source":{"id":"2110.02732","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-10-06T13:27:23Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"a1e3ec8df7eeb2c0d3e45fdd5a6bf5061868a52282691624eac477e30d2fc96b","abstract_canon_sha256":"3498c8682d30c7906e54683cbbc9a735eb1f8a2084034ea4445e2327acdd2a20"},"schema_version":"1.0"},"canonical_sha256":"dbf55823adcd12fa86b0921efbaa3ea7e7e6e626442fae486fbdb01a3ca7a69a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:02:23.602764Z","signature_b64":"9f79oNj93WcOIzPoZKuHvczkeNAjHgGTWpRWjVIuxQF/CnXnKb6LA9bhtF1n3kvi53IA4YdbKxiysyTIpPTxBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dbf55823adcd12fa86b0921efbaa3ea7e7e6e626442fae486fbdb01a3ca7a69a","last_reissued_at":"2026-07-05T05:02:23.602355Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:02:23.602355Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2110.02732","source_version":4,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T05:02:23Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Cld1pJkQbt4qhd3GSVlpiVlkSuESzyULFkY4ZRf7ta5ugyRlGT/utid8z0ElFZLfm/1nWLn4d48irBJLSYUvAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T19:42:38.360152Z"},"content_sha256":"6d6f6228df77bd49f49e14266a943957b26060052989fba5010ba34da8a1230d","schema_version":"1.0","event_id":"sha256:6d6f6228df77bd49f49e14266a943957b26060052989fba5010ba34da8a1230d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2021:3P2VQI5NZUJPVBVQSIPPXKR6U7","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"On Margin Maximization in Linear and ReLU Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Gal Vardi, Nathan Srebro, Ohad Shamir","submitted_at":"2021-10-06T13:27:23Z","abstract_excerpt":"The implicit bias of neural networks has been extensively studied in recent years. Lyu and Li [2019] showed that in homogeneous networks trained with the exponential or the logistic loss, gradient flow converges to a KKT point of the max margin problem in the parameter space. However, that leaves open the question of whether this point will generally be an actual optimum of the max margin problem. In this paper, we study this question in detail, for several neural network architectures involving linear and ReLU activations. Perhaps surprisingly, we show that in many cases, the KKT point is not"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2110.02732","kind":"arxiv","version":4},"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/2110.02732/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":"2026-07-05T05:02:23Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FuJl48UHkigUj3MJyq/hCepExGD+LGwTCibSQacl6JOAsnNHK3UZwE6THNrbamIhP7pFpJa4Cqm9VZv74wTsAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-10T19:42:38.360721Z"},"content_sha256":"c141834277165500c797de97fb2519d274cfa5c855740aa7966e0b316bbc19bf","schema_version":"1.0","event_id":"sha256:c141834277165500c797de97fb2519d274cfa5c855740aa7966e0b316bbc19bf"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/3P2VQI5NZUJPVBVQSIPPXKR6U7/bundle.json","state_url":"https://pith.science/pith/3P2VQI5NZUJPVBVQSIPPXKR6U7/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/3P2VQI5NZUJPVBVQSIPPXKR6U7/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":"2026-08-10T19:42:38Z","links":{"resolver":"https://pith.science/pith/3P2VQI5NZUJPVBVQSIPPXKR6U7","bundle":"https://pith.science/pith/3P2VQI5NZUJPVBVQSIPPXKR6U7/bundle.json","state":"https://pith.science/pith/3P2VQI5NZUJPVBVQSIPPXKR6U7/state.json","well_known_bundle":"https://pith.science/.well-known/pith/3P2VQI5NZUJPVBVQSIPPXKR6U7/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2021:3P2VQI5NZUJPVBVQSIPPXKR6U7","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":"3498c8682d30c7906e54683cbbc9a735eb1f8a2084034ea4445e2327acdd2a20","cross_cats_sorted":["stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-10-06T13:27:23Z","title_canon_sha256":"a1e3ec8df7eeb2c0d3e45fdd5a6bf5061868a52282691624eac477e30d2fc96b"},"schema_version":"1.0","source":{"id":"2110.02732","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2110.02732","created_at":"2026-07-05T05:02:23Z"},{"alias_kind":"arxiv_version","alias_value":"2110.02732v4","created_at":"2026-07-05T05:02:23Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2110.02732","created_at":"2026-07-05T05:02:23Z"},{"alias_kind":"pith_short_12","alias_value":"3P2VQI5NZUJP","created_at":"2026-07-05T05:02:23Z"},{"alias_kind":"pith_short_16","alias_value":"3P2VQI5NZUJPVBVQ","created_at":"2026-07-05T05:02:23Z"},{"alias_kind":"pith_short_8","alias_value":"3P2VQI5N","created_at":"2026-07-05T05:02:23Z"}],"graph_snapshots":[{"event_id":"sha256:c141834277165500c797de97fb2519d274cfa5c855740aa7966e0b316bbc19bf","target":"graph","created_at":"2026-07-05T05:02:23Z","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/2110.02732/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The implicit bias of neural networks has been extensively studied in recent years. Lyu and Li [2019] showed that in homogeneous networks trained with the exponential or the logistic loss, gradient flow converges to a KKT point of the max margin problem in the parameter space. However, that leaves open the question of whether this point will generally be an actual optimum of the max margin problem. In this paper, we study this question in detail, for several neural network architectures involving linear and ReLU activations. Perhaps surprisingly, we show that in many cases, the KKT point is not","authors_text":"Gal Vardi, Nathan Srebro, Ohad Shamir","cross_cats":["stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-10-06T13:27:23Z","title":"On Margin Maximization in Linear and ReLU Networks"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2110.02732","kind":"arxiv","version":4},"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":"sha256:6d6f6228df77bd49f49e14266a943957b26060052989fba5010ba34da8a1230d","target":"record","created_at":"2026-07-05T05:02:23Z","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":"3498c8682d30c7906e54683cbbc9a735eb1f8a2084034ea4445e2327acdd2a20","cross_cats_sorted":["stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-10-06T13:27:23Z","title_canon_sha256":"a1e3ec8df7eeb2c0d3e45fdd5a6bf5061868a52282691624eac477e30d2fc96b"},"schema_version":"1.0","source":{"id":"2110.02732","kind":"arxiv","version":4}},"canonical_sha256":"dbf55823adcd12fa86b0921efbaa3ea7e7e6e626442fae486fbdb01a3ca7a69a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"dbf55823adcd12fa86b0921efbaa3ea7e7e6e626442fae486fbdb01a3ca7a69a","first_computed_at":"2026-07-05T05:02:23.602355Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T05:02:23.602355Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"9f79oNj93WcOIzPoZKuHvczkeNAjHgGTWpRWjVIuxQF/CnXnKb6LA9bhtF1n3kvi53IA4YdbKxiysyTIpPTxBQ==","signature_status":"signed_v1","signed_at":"2026-07-05T05:02:23.602764Z","signed_message":"canonical_sha256_bytes"},"source_id":"2110.02732","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:6d6f6228df77bd49f49e14266a943957b26060052989fba5010ba34da8a1230d","sha256:c141834277165500c797de97fb2519d274cfa5c855740aa7966e0b316bbc19bf"],"state_sha256":"ef986fc9647cef90cbd0da0acb7e40ee96fb1713c318d06e1bf31683e47a5494"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"TAf/83eRZ27+CqbaECVaXKCx61TrikHxsfjPa7sV0/eN/BlsNgVL2F0NCcy02+c3+Ctqc/qu2l5eiaki2qdZDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-10T19:42:38.365264Z","bundle_sha256":"7e0dae322038573c8e2e4a37e95485c4699fdc88a76fed761fdb890c936c1aa0"}}