{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:L7KE2CCK3QKMYUKDMV6OFJ3MAP","short_pith_number":"pith:L7KE2CCK","schema_version":"1.0","canonical_sha256":"5fd44d084adc14cc5143657ce2a76c03c1d4914e12daa5fc850582577b692661","source":{"kind":"arxiv","id":"2003.13964","version":2},"attestation_state":"computed","paper":{"title":"Regularizing Class-wise Predictions via Self-knowledge Distillation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jinwoo Shin, Jongjin Park, Kimin Lee, Sukmin Yun","submitted_at":"2020-03-31T06:03:51Z","abstract_excerpt":"Deep neural networks with millions of parameters may suffer from poor generalization due to overfitting. To mitigate the issue, we propose a new regularization method that penalizes the predictive distribution between similar samples. In particular, we distill the predictive distribution between different samples of the same label during training. This results in regularizing the dark knowledge (i.e., the knowledge on wrong predictions) of a single network (i.e., a self-knowledge distillation) by forcing it to produce more meaningful and consistent predictions in a class-wise manner. Consequen"},"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":"2003.13964","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-03-31T06:03:51Z","cross_cats_sorted":["cs.CV","stat.ML"],"title_canon_sha256":"7840818f06630bcbddbfe2544f17a3ccfd99805fdd11805be58c2855c465df50","abstract_canon_sha256":"ba1f6b0024552566107043820e5842c1430af4c7d7a7f8389ee38018914d7460"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:53:20.624012Z","signature_b64":"UCg4zIVHr417iaT3tfY92Kf/wjZnn6DB0boxVaJkW/a2TNvkVFnfeCLLoveNQc78AFZdaneNVzjuIL2yKpt4BA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5fd44d084adc14cc5143657ce2a76c03c1d4914e12daa5fc850582577b692661","last_reissued_at":"2026-07-05T00:53:20.623333Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:53:20.623333Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Regularizing Class-wise Predictions via Self-knowledge Distillation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Jinwoo Shin, Jongjin Park, Kimin Lee, Sukmin Yun","submitted_at":"2020-03-31T06:03:51Z","abstract_excerpt":"Deep neural networks with millions of parameters may suffer from poor generalization due to overfitting. To mitigate the issue, we propose a new regularization method that penalizes the predictive distribution between similar samples. In particular, we distill the predictive distribution between different samples of the same label during training. This results in regularizing the dark knowledge (i.e., the knowledge on wrong predictions) of a single network (i.e., a self-knowledge distillation) by forcing it to produce more meaningful and consistent predictions in a class-wise manner. Consequen"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2003.13964","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/2003.13964/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":"2003.13964","created_at":"2026-07-05T00:53:20.623410+00:00"},{"alias_kind":"arxiv_version","alias_value":"2003.13964v2","created_at":"2026-07-05T00:53:20.623410+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2003.13964","created_at":"2026-07-05T00:53:20.623410+00:00"},{"alias_kind":"pith_short_12","alias_value":"L7KE2CCK3QKM","created_at":"2026-07-05T00:53:20.623410+00:00"},{"alias_kind":"pith_short_16","alias_value":"L7KE2CCK3QKMYUKD","created_at":"2026-07-05T00:53:20.623410+00:00"},{"alias_kind":"pith_short_8","alias_value":"L7KE2CCK","created_at":"2026-07-05T00:53:20.623410+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.14751","citing_title":"Self Distillation via Iterative Constructive Perturbations","ref_index":23,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/L7KE2CCK3QKMYUKDMV6OFJ3MAP","json":"https://pith.science/pith/L7KE2CCK3QKMYUKDMV6OFJ3MAP.json","graph_json":"https://pith.science/api/pith-number/L7KE2CCK3QKMYUKDMV6OFJ3MAP/graph.json","events_json":"https://pith.science/api/pith-number/L7KE2CCK3QKMYUKDMV6OFJ3MAP/events.json","paper":"https://pith.science/paper/L7KE2CCK"},"agent_actions":{"view_html":"https://pith.science/pith/L7KE2CCK3QKMYUKDMV6OFJ3MAP","download_json":"https://pith.science/pith/L7KE2CCK3QKMYUKDMV6OFJ3MAP.json","view_paper":"https://pith.science/paper/L7KE2CCK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2003.13964&json=true","fetch_graph":"https://pith.science/api/pith-number/L7KE2CCK3QKMYUKDMV6OFJ3MAP/graph.json","fetch_events":"https://pith.science/api/pith-number/L7KE2CCK3QKMYUKDMV6OFJ3MAP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/L7KE2CCK3QKMYUKDMV6OFJ3MAP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/L7KE2CCK3QKMYUKDMV6OFJ3MAP/action/storage_attestation","attest_author":"https://pith.science/pith/L7KE2CCK3QKMYUKDMV6OFJ3MAP/action/author_attestation","sign_citation":"https://pith.science/pith/L7KE2CCK3QKMYUKDMV6OFJ3MAP/action/citation_signature","submit_replication":"https://pith.science/pith/L7KE2CCK3QKMYUKDMV6OFJ3MAP/action/replication_record"}},"created_at":"2026-07-05T00:53:20.623410+00:00","updated_at":"2026-07-05T00:53:20.623410+00:00"}