{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:6EAVFPGHZ7FGE5SVONBTSNQJML","short_pith_number":"pith:6EAVFPGH","schema_version":"1.0","canonical_sha256":"f10152bcc7cfca627655734339360962c3ca728ebdd046f9d64b639f1bcab47a","source":{"kind":"arxiv","id":"2105.10446","version":3},"attestation_state":"computed","paper":{"title":"ReduNet: A White-box Deep Network from the Principle of Maximizing Rate Reduction","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.IT","math.IT","stat.ML"],"primary_cat":"cs.LG","authors_text":"Chong You, Haozhi Qi, John Wright, Kwan Ho Ryan Chan, Yaodong Yu, Yi Ma","submitted_at":"2021-05-21T16:29:57Z","abstract_excerpt":"This work attempts to provide a plausible theoretical framework that aims to interpret modern deep (convolutional) networks from the principles of data compression and discriminative representation. We argue that for high-dimensional multi-class data, the optimal linear discriminative representation maximizes the coding rate difference between the whole dataset and the average of all the subsets. We show that the basic iterative gradient ascent scheme for optimizing the rate reduction objective naturally leads to a multi-layer deep network, named ReduNet, which shares common characteristics of"},"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":"2105.10446","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-05-21T16:29:57Z","cross_cats_sorted":["cs.CV","cs.IT","math.IT","stat.ML"],"title_canon_sha256":"5b0307bd118c9639912245ab648357ee21be479c688601ce1521429f42544403","abstract_canon_sha256":"6db9e472ff9695631757836ad87714b4e171a7ffe14f2b6d4b4591d8172d69e5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:35:28.826632Z","signature_b64":"STK7kiSemCdE7Gard66r+TFVs/lLWmLixqUTi74EP2BcXOQX6QBmhfDszpbsPox5kC9uw8DYPu9cfltrahG3Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f10152bcc7cfca627655734339360962c3ca728ebdd046f9d64b639f1bcab47a","last_reissued_at":"2026-07-05T03:35:28.826143Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:35:28.826143Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ReduNet: A White-box Deep Network from the Principle of Maximizing Rate Reduction","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","cs.IT","math.IT","stat.ML"],"primary_cat":"cs.LG","authors_text":"Chong You, Haozhi Qi, John Wright, Kwan Ho Ryan Chan, Yaodong Yu, Yi Ma","submitted_at":"2021-05-21T16:29:57Z","abstract_excerpt":"This work attempts to provide a plausible theoretical framework that aims to interpret modern deep (convolutional) networks from the principles of data compression and discriminative representation. We argue that for high-dimensional multi-class data, the optimal linear discriminative representation maximizes the coding rate difference between the whole dataset and the average of all the subsets. We show that the basic iterative gradient ascent scheme for optimizing the rate reduction objective naturally leads to a multi-layer deep network, named ReduNet, which shares common characteristics of"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2105.10446","kind":"arxiv","version":3},"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/2105.10446/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":"2105.10446","created_at":"2026-07-05T03:35:28.826202+00:00"},{"alias_kind":"arxiv_version","alias_value":"2105.10446v3","created_at":"2026-07-05T03:35:28.826202+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2105.10446","created_at":"2026-07-05T03:35:28.826202+00:00"},{"alias_kind":"pith_short_12","alias_value":"6EAVFPGHZ7FG","created_at":"2026-07-05T03:35:28.826202+00:00"},{"alias_kind":"pith_short_16","alias_value":"6EAVFPGHZ7FGE5SV","created_at":"2026-07-05T03:35:28.826202+00:00"},{"alias_kind":"pith_short_8","alias_value":"6EAVFPGH","created_at":"2026-07-05T03:35:28.826202+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/6EAVFPGHZ7FGE5SVONBTSNQJML","json":"https://pith.science/pith/6EAVFPGHZ7FGE5SVONBTSNQJML.json","graph_json":"https://pith.science/api/pith-number/6EAVFPGHZ7FGE5SVONBTSNQJML/graph.json","events_json":"https://pith.science/api/pith-number/6EAVFPGHZ7FGE5SVONBTSNQJML/events.json","paper":"https://pith.science/paper/6EAVFPGH"},"agent_actions":{"view_html":"https://pith.science/pith/6EAVFPGHZ7FGE5SVONBTSNQJML","download_json":"https://pith.science/pith/6EAVFPGHZ7FGE5SVONBTSNQJML.json","view_paper":"https://pith.science/paper/6EAVFPGH","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2105.10446&json=true","fetch_graph":"https://pith.science/api/pith-number/6EAVFPGHZ7FGE5SVONBTSNQJML/graph.json","fetch_events":"https://pith.science/api/pith-number/6EAVFPGHZ7FGE5SVONBTSNQJML/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6EAVFPGHZ7FGE5SVONBTSNQJML/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6EAVFPGHZ7FGE5SVONBTSNQJML/action/storage_attestation","attest_author":"https://pith.science/pith/6EAVFPGHZ7FGE5SVONBTSNQJML/action/author_attestation","sign_citation":"https://pith.science/pith/6EAVFPGHZ7FGE5SVONBTSNQJML/action/citation_signature","submit_replication":"https://pith.science/pith/6EAVFPGHZ7FGE5SVONBTSNQJML/action/replication_record"}},"created_at":"2026-07-05T03:35:28.826202+00:00","updated_at":"2026-07-05T03:35:28.826202+00:00"}