{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:BDOOUN6IMQXVW577C4FY6SGR2G","short_pith_number":"pith:BDOOUN6I","schema_version":"1.0","canonical_sha256":"08dcea37c8642f5b77ff170b8f48d1d1a821488135af63cf00ee97d27699c767","source":{"kind":"arxiv","id":"2605.19533","version":1},"attestation_state":"computed","paper":{"title":"Replacement Learning: Training Neural Networks with Fewer Parameters","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Dongzhi Guan, Hengyu Shi, Jiabin Liu, Jiaji Wang, Junhao Su, Peizhe Wang, Tianyang Han, Yuming Zhang","submitted_at":"2026-05-19T08:34:31Z","abstract_excerpt":"End-to-end training with full-depth backpropagation remains the dominant paradigm for optimizing deep neural networks, but its efficiency deteriorates as models grow deeper. Since every block must be executed and differentiated under a single global objective, full-depth BP introduces substantial parameter redundancy, activation-memory cost, and training latency, especially when neighboring layers exhibit highly correlated learning patterns. Directly skipping or removing layers can reduce cost, but often weakens representation capacity or requires architecture-specific reuse designs. In this p"},"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":"2605.19533","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-19T08:34:31Z","cross_cats_sorted":[],"title_canon_sha256":"3c0dafac9ee271029c1607ba7ee5503bfc29f601f8b3114ce4327006d5935436","abstract_canon_sha256":"6f3bff519a3c35bd9af0ff5b913e8398b4efb4fcd7babe02a84a27f1db171acf"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T01:05:50.554536Z","signature_b64":"uN7JOVV9JItAJ4a+i5nrSophZU9uIrzMVBejvr1oM+rxbYLIgmuH9Df6ddNRMPSSnh9jQb1I1v7/0NXjz+gxAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"08dcea37c8642f5b77ff170b8f48d1d1a821488135af63cf00ee97d27699c767","last_reissued_at":"2026-05-20T01:05:50.553692Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T01:05:50.553692Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Replacement Learning: Training Neural Networks with Fewer Parameters","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Dongzhi Guan, Hengyu Shi, Jiabin Liu, Jiaji Wang, Junhao Su, Peizhe Wang, Tianyang Han, Yuming Zhang","submitted_at":"2026-05-19T08:34:31Z","abstract_excerpt":"End-to-end training with full-depth backpropagation remains the dominant paradigm for optimizing deep neural networks, but its efficiency deteriorates as models grow deeper. Since every block must be executed and differentiated under a single global objective, full-depth BP introduces substantial parameter redundancy, activation-memory cost, and training latency, especially when neighboring layers exhibit highly correlated learning patterns. Directly skipping or removing layers can reduce cost, but often weakens representation capacity or requires architecture-specific reuse designs. In this p"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.19533","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/2605.19533/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":"2605.19533","created_at":"2026-05-20T01:05:50.553832+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.19533v1","created_at":"2026-05-20T01:05:50.553832+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.19533","created_at":"2026-05-20T01:05:50.553832+00:00"},{"alias_kind":"pith_short_12","alias_value":"BDOOUN6IMQXV","created_at":"2026-05-20T01:05:50.553832+00:00"},{"alias_kind":"pith_short_16","alias_value":"BDOOUN6IMQXVW577","created_at":"2026-05-20T01:05:50.553832+00:00"},{"alias_kind":"pith_short_8","alias_value":"BDOOUN6I","created_at":"2026-05-20T01:05:50.553832+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/BDOOUN6IMQXVW577C4FY6SGR2G","json":"https://pith.science/pith/BDOOUN6IMQXVW577C4FY6SGR2G.json","graph_json":"https://pith.science/api/pith-number/BDOOUN6IMQXVW577C4FY6SGR2G/graph.json","events_json":"https://pith.science/api/pith-number/BDOOUN6IMQXVW577C4FY6SGR2G/events.json","paper":"https://pith.science/paper/BDOOUN6I"},"agent_actions":{"view_html":"https://pith.science/pith/BDOOUN6IMQXVW577C4FY6SGR2G","download_json":"https://pith.science/pith/BDOOUN6IMQXVW577C4FY6SGR2G.json","view_paper":"https://pith.science/paper/BDOOUN6I","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.19533&json=true","fetch_graph":"https://pith.science/api/pith-number/BDOOUN6IMQXVW577C4FY6SGR2G/graph.json","fetch_events":"https://pith.science/api/pith-number/BDOOUN6IMQXVW577C4FY6SGR2G/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BDOOUN6IMQXVW577C4FY6SGR2G/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BDOOUN6IMQXVW577C4FY6SGR2G/action/storage_attestation","attest_author":"https://pith.science/pith/BDOOUN6IMQXVW577C4FY6SGR2G/action/author_attestation","sign_citation":"https://pith.science/pith/BDOOUN6IMQXVW577C4FY6SGR2G/action/citation_signature","submit_replication":"https://pith.science/pith/BDOOUN6IMQXVW577C4FY6SGR2G/action/replication_record"}},"created_at":"2026-05-20T01:05:50.553832+00:00","updated_at":"2026-05-20T01:05:50.553832+00:00"}