{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:IQCRMFTLRO456QVZ6JJOO7TV32","short_pith_number":"pith:IQCRMFTL","schema_version":"1.0","canonical_sha256":"440516166b8bb9df42b9f252e77e75dead45282856a9e2b132c110bb2c252d7e","source":{"kind":"arxiv","id":"2310.02012","version":1},"attestation_state":"computed","paper":{"title":"Towards Training Without Depth Limits: Batch Normalization Without Gradient Explosion","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Alexander Immer, Alexandru Meterez, Amir Joudaki, Francesco Orabona, Gunnar R\\\"atsch, Hadi Daneshmand","submitted_at":"2023-10-03T12:35:02Z","abstract_excerpt":"Normalization layers are one of the key building blocks for deep neural networks. Several theoretical studies have shown that batch normalization improves the signal propagation, by avoiding the representations from becoming collinear across the layers. However, results on mean-field theory of batch normalization also conclude that this benefit comes at the expense of exploding gradients in depth. Motivated by these two aspects of batch normalization, in this study we pose the following question: \"Can a batch-normalized network keep the optimal signal propagation properties, but avoid explodin"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2310.02012","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-10-03T12:35:02Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"7c64368c136d37ae85ff2dedc49d3df478b5b606b44c50f2feccf50db618fc2b","abstract_canon_sha256":"631ef34af7c1c1ae85611ebda6401a9597b8568cac6e0c2bd6e1b9e35bcc6d11"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:56:48.966503Z","signature_b64":"dA50mlZN4Ix9iFDexlQ8835MQIKwR2KwMfqUSgShn8xqZrGvXLmfsVXge/D1SEabSKaQrEXL1821ZSMLC0sKDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"440516166b8bb9df42b9f252e77e75dead45282856a9e2b132c110bb2c252d7e","last_reissued_at":"2026-07-05T06:56:48.966064Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:56:48.966064Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Towards Training Without Depth Limits: Batch Normalization Without Gradient Explosion","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Alexander Immer, Alexandru Meterez, Amir Joudaki, Francesco Orabona, Gunnar R\\\"atsch, Hadi Daneshmand","submitted_at":"2023-10-03T12:35:02Z","abstract_excerpt":"Normalization layers are one of the key building blocks for deep neural networks. Several theoretical studies have shown that batch normalization improves the signal propagation, by avoiding the representations from becoming collinear across the layers. However, results on mean-field theory of batch normalization also conclude that this benefit comes at the expense of exploding gradients in depth. Motivated by these two aspects of batch normalization, in this study we pose the following question: \"Can a batch-normalized network keep the optimal signal propagation properties, but avoid explodin"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.02012","kind":"arxiv","version":1},"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/2310.02012/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2310.02012","created_at":"2026-07-05T06:56:48.966128+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.02012v1","created_at":"2026-07-05T06:56:48.966128+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.02012","created_at":"2026-07-05T06:56:48.966128+00:00"},{"alias_kind":"pith_short_12","alias_value":"IQCRMFTLRO45","created_at":"2026-07-05T06:56:48.966128+00:00"},{"alias_kind":"pith_short_16","alias_value":"IQCRMFTLRO456QVZ","created_at":"2026-07-05T06:56:48.966128+00:00"},{"alias_kind":"pith_short_8","alias_value":"IQCRMFTL","created_at":"2026-07-05T06:56:48.966128+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IQCRMFTLRO456QVZ6JJOO7TV32","json":"https://pith.science/pith/IQCRMFTLRO456QVZ6JJOO7TV32.json","graph_json":"https://pith.science/api/pith-number/IQCRMFTLRO456QVZ6JJOO7TV32/graph.json","events_json":"https://pith.science/api/pith-number/IQCRMFTLRO456QVZ6JJOO7TV32/events.json","paper":"https://pith.science/paper/IQCRMFTL"},"agent_actions":{"view_html":"https://pith.science/pith/IQCRMFTLRO456QVZ6JJOO7TV32","download_json":"https://pith.science/pith/IQCRMFTLRO456QVZ6JJOO7TV32.json","view_paper":"https://pith.science/paper/IQCRMFTL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.02012&json=true","fetch_graph":"https://pith.science/api/pith-number/IQCRMFTLRO456QVZ6JJOO7TV32/graph.json","fetch_events":"https://pith.science/api/pith-number/IQCRMFTLRO456QVZ6JJOO7TV32/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IQCRMFTLRO456QVZ6JJOO7TV32/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IQCRMFTLRO456QVZ6JJOO7TV32/action/storage_attestation","attest_author":"https://pith.science/pith/IQCRMFTLRO456QVZ6JJOO7TV32/action/author_attestation","sign_citation":"https://pith.science/pith/IQCRMFTLRO456QVZ6JJOO7TV32/action/citation_signature","submit_replication":"https://pith.science/pith/IQCRMFTLRO456QVZ6JJOO7TV32/action/replication_record"}},"created_at":"2026-07-05T06:56:48.966128+00:00","updated_at":"2026-07-05T06:56:48.966128+00:00"}