{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:ZCDKP3OF3FACQH75FG7VY4C2TC","short_pith_number":"pith:ZCDKP3OF","schema_version":"1.0","canonical_sha256":"c886a7edc5d940281ffd29bf5c705a98a95ab2b1d9124652b9e893754f567a5b","source":{"kind":"arxiv","id":"2303.02733","version":2},"attestation_state":"computed","paper":{"title":"Reparameterization through Spatial Gradient Scaling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.LG","authors_text":"Alexander Detkov, Di Niu, Jialin Zhang, Mohammad Salameh, Muhammad Fetrat Qharabagh, Shangling Jui, Wei Lui","submitted_at":"2023-03-05T17:57:33Z","abstract_excerpt":"Reparameterization aims to improve the generalization of deep neural networks by transforming convolutional layers into equivalent multi-branched structures during training. However, there exists a gap in understanding how reparameterization may change and benefit the learning process of neural networks. In this paper, we present a novel spatial gradient scaling method to redistribute learning focus among weights in convolutional networks. We prove that spatial gradient scaling achieves the same learning dynamics as a branched reparameterization yet without introducing structural changes into "},"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":"2303.02733","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-03-05T17:57:33Z","cross_cats_sorted":["cs.AI","cs.CV"],"title_canon_sha256":"dd087ff337678a2932bfcc946bdebf80d7d8ccef8875608c6288e3f089202518","abstract_canon_sha256":"540c90a7bbfccb11ee65912364433f3d30124c5c370623cd53d3b822e61d7309"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:48:53.925110Z","signature_b64":"rxZClMXmRc5AQ68Xg4s5RkRxLXAFwOpfn3QqMdFBQhBiTLGhnFdOPIS0L38A3LyQ+gHKzy2vJVCsagZMqkNMCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c886a7edc5d940281ffd29bf5c705a98a95ab2b1d9124652b9e893754f567a5b","last_reissued_at":"2026-07-05T05:48:53.924601Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:48:53.924601Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Reparameterization through Spatial Gradient Scaling","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.LG","authors_text":"Alexander Detkov, Di Niu, Jialin Zhang, Mohammad Salameh, Muhammad Fetrat Qharabagh, Shangling Jui, Wei Lui","submitted_at":"2023-03-05T17:57:33Z","abstract_excerpt":"Reparameterization aims to improve the generalization of deep neural networks by transforming convolutional layers into equivalent multi-branched structures during training. However, there exists a gap in understanding how reparameterization may change and benefit the learning process of neural networks. In this paper, we present a novel spatial gradient scaling method to redistribute learning focus among weights in convolutional networks. We prove that spatial gradient scaling achieves the same learning dynamics as a branched reparameterization yet without introducing structural changes into "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2303.02733","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/2303.02733/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":"2303.02733","created_at":"2026-07-05T05:48:53.924663+00:00"},{"alias_kind":"arxiv_version","alias_value":"2303.02733v2","created_at":"2026-07-05T05:48:53.924663+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2303.02733","created_at":"2026-07-05T05:48:53.924663+00:00"},{"alias_kind":"pith_short_12","alias_value":"ZCDKP3OF3FAC","created_at":"2026-07-05T05:48:53.924663+00:00"},{"alias_kind":"pith_short_16","alias_value":"ZCDKP3OF3FACQH75","created_at":"2026-07-05T05:48:53.924663+00:00"},{"alias_kind":"pith_short_8","alias_value":"ZCDKP3OF","created_at":"2026-07-05T05:48:53.924663+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/ZCDKP3OF3FACQH75FG7VY4C2TC","json":"https://pith.science/pith/ZCDKP3OF3FACQH75FG7VY4C2TC.json","graph_json":"https://pith.science/api/pith-number/ZCDKP3OF3FACQH75FG7VY4C2TC/graph.json","events_json":"https://pith.science/api/pith-number/ZCDKP3OF3FACQH75FG7VY4C2TC/events.json","paper":"https://pith.science/paper/ZCDKP3OF"},"agent_actions":{"view_html":"https://pith.science/pith/ZCDKP3OF3FACQH75FG7VY4C2TC","download_json":"https://pith.science/pith/ZCDKP3OF3FACQH75FG7VY4C2TC.json","view_paper":"https://pith.science/paper/ZCDKP3OF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2303.02733&json=true","fetch_graph":"https://pith.science/api/pith-number/ZCDKP3OF3FACQH75FG7VY4C2TC/graph.json","fetch_events":"https://pith.science/api/pith-number/ZCDKP3OF3FACQH75FG7VY4C2TC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ZCDKP3OF3FACQH75FG7VY4C2TC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ZCDKP3OF3FACQH75FG7VY4C2TC/action/storage_attestation","attest_author":"https://pith.science/pith/ZCDKP3OF3FACQH75FG7VY4C2TC/action/author_attestation","sign_citation":"https://pith.science/pith/ZCDKP3OF3FACQH75FG7VY4C2TC/action/citation_signature","submit_replication":"https://pith.science/pith/ZCDKP3OF3FACQH75FG7VY4C2TC/action/replication_record"}},"created_at":"2026-07-05T05:48:53.924663+00:00","updated_at":"2026-07-05T05:48:53.924663+00:00"}