{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:ISLXRVJA3GUTMPE4NDWA3SI4TH","short_pith_number":"pith:ISLXRVJA","canonical_record":{"source":{"id":"2603.04895","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2026-03-05T07:36:07Z","cross_cats_sorted":["cs.LG","math.OC"],"title_canon_sha256":"a00a360eeefd0bcab8d04a986fb1cad5a12cafea31fed4c425e7d3db733467ab","abstract_canon_sha256":"20bfcfed2341db091932c19d94c0109ca845ef032fa17c6198bfca4002bde3d4"},"schema_version":"1.0"},"canonical_sha256":"449778d520d9a9363c9c68ec0dc91c99d4a8148ca3e4e552e92fe43376c73825","source":{"kind":"arxiv","id":"2603.04895","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2603.04895","created_at":"2026-06-19T16:11:21Z"},{"alias_kind":"arxiv_version","alias_value":"2603.04895v2","created_at":"2026-06-19T16:11:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.04895","created_at":"2026-06-19T16:11:21Z"},{"alias_kind":"pith_short_12","alias_value":"ISLXRVJA3GUT","created_at":"2026-06-19T16:11:21Z"},{"alias_kind":"pith_short_16","alias_value":"ISLXRVJA3GUTMPE4","created_at":"2026-06-19T16:11:21Z"},{"alias_kind":"pith_short_8","alias_value":"ISLXRVJA","created_at":"2026-06-19T16:11:21Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:ISLXRVJA3GUTMPE4NDWA3SI4TH","target":"record","payload":{"canonical_record":{"source":{"id":"2603.04895","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2026-03-05T07:36:07Z","cross_cats_sorted":["cs.LG","math.OC"],"title_canon_sha256":"a00a360eeefd0bcab8d04a986fb1cad5a12cafea31fed4c425e7d3db733467ab","abstract_canon_sha256":"20bfcfed2341db091932c19d94c0109ca845ef032fa17c6198bfca4002bde3d4"},"schema_version":"1.0"},"canonical_sha256":"449778d520d9a9363c9c68ec0dc91c99d4a8148ca3e4e552e92fe43376c73825","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-19T16:11:21.574181Z","signature_b64":"dolZi0TJusN6aksSqXrPBMPyfzX1uX8n0X0g9fo29K6Binn/gWxSIKbndWBHMtSz3Hp4pXO5KMqj2GOoFbJ/AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"449778d520d9a9363c9c68ec0dc91c99d4a8148ca3e4e552e92fe43376c73825","last_reissued_at":"2026-06-19T16:11:21.573729Z","signature_status":"signed_v1","first_computed_at":"2026-06-19T16:11:21.573729Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2603.04895","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":"2026-06-19T16:11:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5/2S5qGehY6eBZG+bsBgeLrC2jxdAmMZGFVPCe4JSFTnTCFnlr3Rz14e6P3iR2kFX4ch3KupkRfOknwXhXlWDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-01T13:43:54.130743Z"},"content_sha256":"f40a9850a673932abb83215de61808648aab46494c58d35e834aa459433719a1","schema_version":"1.0","event_id":"sha256:f40a9850a673932abb83215de61808648aab46494c58d35e834aa459433719a1"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:ISLXRVJA3GUTMPE4NDWA3SI4TH","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"How Does the ReLU Activation Affect the Implicit Bias of Gradient Descent on High-dimensional Neural Network Regression?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","math.OC"],"primary_cat":"stat.ML","authors_text":"Guanghui Wang, Kuo-Wei Lai, Molei Tao, Vidya Muthukumar","submitted_at":"2026-03-05T07:36:07Z","abstract_excerpt":"Overparameterized ML models, including neural networks, typically induce underdetermined training objectives with multiple global minima. The implicit bias refers to the limiting global minimum that is attained by a common optimization algorithm, such as gradient descent (GD). In this paper, we characterize the implicit bias of GD for training a shallow ReLU model with the squared loss on high-dimensional random features. Prior work (Vardi and Shamir, 2021) showed that the implicit bias does not exist in the worst-case, or corresponds exactly to the minimum-$\\ell_2$-norm interpolating solution"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.04895","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/2603.04895/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":"2026-06-19T16:11:21Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"6VoHjJQi2J7SALLL2wn1aLZC3cpzODm4FAprIUxOINBPoQ8IKh2Hy+kLW379ENjvl5KxzuyK9lTgowBJptd/Cg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-01T13:43:54.131428Z"},"content_sha256":"f1122e4b4027d830008c64ff24ce57dca3404e5e03954089afa3b536d347c20b","schema_version":"1.0","event_id":"sha256:f1122e4b4027d830008c64ff24ce57dca3404e5e03954089afa3b536d347c20b"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ISLXRVJA3GUTMPE4NDWA3SI4TH/bundle.json","state_url":"https://pith.science/pith/ISLXRVJA3GUTMPE4NDWA3SI4TH/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ISLXRVJA3GUTMPE4NDWA3SI4TH/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":"2026-07-01T13:43:54Z","links":{"resolver":"https://pith.science/pith/ISLXRVJA3GUTMPE4NDWA3SI4TH","bundle":"https://pith.science/pith/ISLXRVJA3GUTMPE4NDWA3SI4TH/bundle.json","state":"https://pith.science/pith/ISLXRVJA3GUTMPE4NDWA3SI4TH/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ISLXRVJA3GUTMPE4NDWA3SI4TH/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:ISLXRVJA3GUTMPE4NDWA3SI4TH","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":"20bfcfed2341db091932c19d94c0109ca845ef032fa17c6198bfca4002bde3d4","cross_cats_sorted":["cs.LG","math.OC"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2026-03-05T07:36:07Z","title_canon_sha256":"a00a360eeefd0bcab8d04a986fb1cad5a12cafea31fed4c425e7d3db733467ab"},"schema_version":"1.0","source":{"id":"2603.04895","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2603.04895","created_at":"2026-06-19T16:11:21Z"},{"alias_kind":"arxiv_version","alias_value":"2603.04895v2","created_at":"2026-06-19T16:11:21Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2603.04895","created_at":"2026-06-19T16:11:21Z"},{"alias_kind":"pith_short_12","alias_value":"ISLXRVJA3GUT","created_at":"2026-06-19T16:11:21Z"},{"alias_kind":"pith_short_16","alias_value":"ISLXRVJA3GUTMPE4","created_at":"2026-06-19T16:11:21Z"},{"alias_kind":"pith_short_8","alias_value":"ISLXRVJA","created_at":"2026-06-19T16:11:21Z"}],"graph_snapshots":[{"event_id":"sha256:f1122e4b4027d830008c64ff24ce57dca3404e5e03954089afa3b536d347c20b","target":"graph","created_at":"2026-06-19T16:11:21Z","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/2603.04895/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Overparameterized ML models, including neural networks, typically induce underdetermined training objectives with multiple global minima. The implicit bias refers to the limiting global minimum that is attained by a common optimization algorithm, such as gradient descent (GD). In this paper, we characterize the implicit bias of GD for training a shallow ReLU model with the squared loss on high-dimensional random features. Prior work (Vardi and Shamir, 2021) showed that the implicit bias does not exist in the worst-case, or corresponds exactly to the minimum-$\\ell_2$-norm interpolating solution","authors_text":"Guanghui Wang, Kuo-Wei Lai, Molei Tao, Vidya Muthukumar","cross_cats":["cs.LG","math.OC"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2026-03-05T07:36:07Z","title":"How Does the ReLU Activation Affect the Implicit Bias of Gradient Descent on High-dimensional Neural Network Regression?"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2603.04895","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:f40a9850a673932abb83215de61808648aab46494c58d35e834aa459433719a1","target":"record","created_at":"2026-06-19T16:11:21Z","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":"20bfcfed2341db091932c19d94c0109ca845ef032fa17c6198bfca4002bde3d4","cross_cats_sorted":["cs.LG","math.OC"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2026-03-05T07:36:07Z","title_canon_sha256":"a00a360eeefd0bcab8d04a986fb1cad5a12cafea31fed4c425e7d3db733467ab"},"schema_version":"1.0","source":{"id":"2603.04895","kind":"arxiv","version":2}},"canonical_sha256":"449778d520d9a9363c9c68ec0dc91c99d4a8148ca3e4e552e92fe43376c73825","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"449778d520d9a9363c9c68ec0dc91c99d4a8148ca3e4e552e92fe43376c73825","first_computed_at":"2026-06-19T16:11:21.573729Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-19T16:11:21.573729Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"dolZi0TJusN6aksSqXrPBMPyfzX1uX8n0X0g9fo29K6Binn/gWxSIKbndWBHMtSz3Hp4pXO5KMqj2GOoFbJ/AA==","signature_status":"signed_v1","signed_at":"2026-06-19T16:11:21.574181Z","signed_message":"canonical_sha256_bytes"},"source_id":"2603.04895","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f40a9850a673932abb83215de61808648aab46494c58d35e834aa459433719a1","sha256:f1122e4b4027d830008c64ff24ce57dca3404e5e03954089afa3b536d347c20b"],"state_sha256":"aaf2d0f77170cc115418b1c75ce3f42b6b093f30c87dc56e4d6164208c273def"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rEpfnMSDh34cBGwd47R8+F12/FxhzkXk7t3B3+5DOUS3tPsknVapLVQkSs7he0yddoMExGf/6W1VLJ2cRqS0CA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-01T13:43:54.134472Z","bundle_sha256":"1d984572eb5df1090e0ceeb19a72af03014772be48894a4d4415318fec042124"}}