{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:FLVK7VXV5Y5RZ5EBZEDRR2RC5A","short_pith_number":"pith:FLVK7VXV","schema_version":"1.0","canonical_sha256":"2aeaafd6f5ee3b1cf481c90718ea22e82d5bc36e3e4f87a85c8da1fcd22ce8b4","source":{"kind":"arxiv","id":"2112.05958","version":4},"attestation_state":"computed","paper":{"title":"You Only Need End-to-End Training for Long-Tailed Recognition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Zhiwei Zhang","submitted_at":"2021-12-11T11:44:09Z","abstract_excerpt":"The generalization gap on the long-tailed data sets is largely owing to most categories only occupying a few training samples. Decoupled training achieves better performance by training backbone and classifier separately. What causes the poorer performance of end-to-end model training (e.g., logits margin-based methods)? In this work, we identify a key factor that affects the learning of the classifier: the channel-correlated features with low entropy before inputting into the classifier. From the perspective of information theory, we analyze why cross-entropy loss tends to produce highly corr"},"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":"2112.05958","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2021-12-11T11:44:09Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"d2a6577f4a847ee4d7f657fd93afb7da8f50538b0e8729363fa21dd3b6263a39","abstract_canon_sha256":"456091e441e26e856467bc005dd79035deae5d3eac0693fa962b5a5447e212a3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:49:52.774903Z","signature_b64":"ekBAUVBsgzqQRIJgBmsGIxbCLunDlql/dGXWG6c7FxJ1yzHU+3pIj/KBgTUIrNM4tHHcHU900OC8pxfa9dIvCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2aeaafd6f5ee3b1cf481c90718ea22e82d5bc36e3e4f87a85c8da1fcd22ce8b4","last_reissued_at":"2026-07-05T05:49:52.774394Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:49:52.774394Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"You Only Need End-to-End Training for Long-Tailed Recognition","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CV","authors_text":"Zhiwei Zhang","submitted_at":"2021-12-11T11:44:09Z","abstract_excerpt":"The generalization gap on the long-tailed data sets is largely owing to most categories only occupying a few training samples. Decoupled training achieves better performance by training backbone and classifier separately. What causes the poorer performance of end-to-end model training (e.g., logits margin-based methods)? In this work, we identify a key factor that affects the learning of the classifier: the channel-correlated features with low entropy before inputting into the classifier. From the perspective of information theory, we analyze why cross-entropy loss tends to produce highly corr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2112.05958","kind":"arxiv","version":4},"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/2112.05958/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":"2112.05958","created_at":"2026-07-05T05:49:52.774451+00:00"},{"alias_kind":"arxiv_version","alias_value":"2112.05958v4","created_at":"2026-07-05T05:49:52.774451+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2112.05958","created_at":"2026-07-05T05:49:52.774451+00:00"},{"alias_kind":"pith_short_12","alias_value":"FLVK7VXV5Y5R","created_at":"2026-07-05T05:49:52.774451+00:00"},{"alias_kind":"pith_short_16","alias_value":"FLVK7VXV5Y5RZ5EB","created_at":"2026-07-05T05:49:52.774451+00:00"},{"alias_kind":"pith_short_8","alias_value":"FLVK7VXV","created_at":"2026-07-05T05:49:52.774451+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/FLVK7VXV5Y5RZ5EBZEDRR2RC5A","json":"https://pith.science/pith/FLVK7VXV5Y5RZ5EBZEDRR2RC5A.json","graph_json":"https://pith.science/api/pith-number/FLVK7VXV5Y5RZ5EBZEDRR2RC5A/graph.json","events_json":"https://pith.science/api/pith-number/FLVK7VXV5Y5RZ5EBZEDRR2RC5A/events.json","paper":"https://pith.science/paper/FLVK7VXV"},"agent_actions":{"view_html":"https://pith.science/pith/FLVK7VXV5Y5RZ5EBZEDRR2RC5A","download_json":"https://pith.science/pith/FLVK7VXV5Y5RZ5EBZEDRR2RC5A.json","view_paper":"https://pith.science/paper/FLVK7VXV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2112.05958&json=true","fetch_graph":"https://pith.science/api/pith-number/FLVK7VXV5Y5RZ5EBZEDRR2RC5A/graph.json","fetch_events":"https://pith.science/api/pith-number/FLVK7VXV5Y5RZ5EBZEDRR2RC5A/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FLVK7VXV5Y5RZ5EBZEDRR2RC5A/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FLVK7VXV5Y5RZ5EBZEDRR2RC5A/action/storage_attestation","attest_author":"https://pith.science/pith/FLVK7VXV5Y5RZ5EBZEDRR2RC5A/action/author_attestation","sign_citation":"https://pith.science/pith/FLVK7VXV5Y5RZ5EBZEDRR2RC5A/action/citation_signature","submit_replication":"https://pith.science/pith/FLVK7VXV5Y5RZ5EBZEDRR2RC5A/action/replication_record"}},"created_at":"2026-07-05T05:49:52.774451+00:00","updated_at":"2026-07-05T05:49:52.774451+00:00"}