{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:A2QH4JN5CKTN2L2K3V4ZKPLURF","short_pith_number":"pith:A2QH4JN5","schema_version":"1.0","canonical_sha256":"06a07e25bd12a6dd2f4add79953d748951b0e56d46958e66642f5c6dc0fd6b8e","source":{"kind":"arxiv","id":"2301.13304","version":1},"attestation_state":"computed","paper":{"title":"Understanding Self-Distillation in the Presence of Label Noise","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Rudrajit Das, Sujay Sanghavi","submitted_at":"2023-01-30T21:50:30Z","abstract_excerpt":"Self-distillation (SD) is the process of first training a \\enquote{teacher} model and then using its predictions to train a \\enquote{student} model with the \\textit{same} architecture. Specifically, the student's objective function is $\\big(\\xi*\\ell(\\text{teacher's predictions}, \\text{ student's predictions}) + (1-\\xi)*\\ell(\\text{given labels}, \\text{ student's predictions})\\big)$, where $\\ell$ is some loss function and $\\xi$ is some parameter $\\in [0,1]$. Empirically, SD has been observed to provide performance gains in several settings. In this paper, we theoretically characterize the effect"},"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":"2301.13304","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2023-01-30T21:50:30Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"4a1a0a845dd891fcbb65c19bab8ec9beb78a419852678fbb9f1964fc203f1357","abstract_canon_sha256":"95b2b98b4fc3e9a1864e5cbc5403462a06cc773eee4dad9caa6a2d820c655b71"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T05:37:13.521091Z","signature_b64":"PELx4t8G9VuRcnJDeHr/4RThpq6d9eP0QoJH1WIkzfZqPJhfQtePL1CD8XVAbxCI82dxyqvzjL+koKAoLq9jBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"06a07e25bd12a6dd2f4add79953d748951b0e56d46958e66642f5c6dc0fd6b8e","last_reissued_at":"2026-07-05T05:37:13.520586Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T05:37:13.520586Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Understanding Self-Distillation in the Presence of Label Noise","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Rudrajit Das, Sujay Sanghavi","submitted_at":"2023-01-30T21:50:30Z","abstract_excerpt":"Self-distillation (SD) is the process of first training a \\enquote{teacher} model and then using its predictions to train a \\enquote{student} model with the \\textit{same} architecture. Specifically, the student's objective function is $\\big(\\xi*\\ell(\\text{teacher's predictions}, \\text{ student's predictions}) + (1-\\xi)*\\ell(\\text{given labels}, \\text{ student's predictions})\\big)$, where $\\ell$ is some loss function and $\\xi$ is some parameter $\\in [0,1]$. Empirically, SD has been observed to provide performance gains in several settings. In this paper, we theoretically characterize the effect"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2301.13304","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/2301.13304/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":"2301.13304","created_at":"2026-07-05T05:37:13.520643+00:00"},{"alias_kind":"arxiv_version","alias_value":"2301.13304v1","created_at":"2026-07-05T05:37:13.520643+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2301.13304","created_at":"2026-07-05T05:37:13.520643+00:00"},{"alias_kind":"pith_short_12","alias_value":"A2QH4JN5CKTN","created_at":"2026-07-05T05:37:13.520643+00:00"},{"alias_kind":"pith_short_16","alias_value":"A2QH4JN5CKTN2L2K","created_at":"2026-07-05T05:37:13.520643+00:00"},{"alias_kind":"pith_short_8","alias_value":"A2QH4JN5","created_at":"2026-07-05T05:37:13.520643+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/A2QH4JN5CKTN2L2K3V4ZKPLURF","json":"https://pith.science/pith/A2QH4JN5CKTN2L2K3V4ZKPLURF.json","graph_json":"https://pith.science/api/pith-number/A2QH4JN5CKTN2L2K3V4ZKPLURF/graph.json","events_json":"https://pith.science/api/pith-number/A2QH4JN5CKTN2L2K3V4ZKPLURF/events.json","paper":"https://pith.science/paper/A2QH4JN5"},"agent_actions":{"view_html":"https://pith.science/pith/A2QH4JN5CKTN2L2K3V4ZKPLURF","download_json":"https://pith.science/pith/A2QH4JN5CKTN2L2K3V4ZKPLURF.json","view_paper":"https://pith.science/paper/A2QH4JN5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2301.13304&json=true","fetch_graph":"https://pith.science/api/pith-number/A2QH4JN5CKTN2L2K3V4ZKPLURF/graph.json","fetch_events":"https://pith.science/api/pith-number/A2QH4JN5CKTN2L2K3V4ZKPLURF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/A2QH4JN5CKTN2L2K3V4ZKPLURF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/A2QH4JN5CKTN2L2K3V4ZKPLURF/action/storage_attestation","attest_author":"https://pith.science/pith/A2QH4JN5CKTN2L2K3V4ZKPLURF/action/author_attestation","sign_citation":"https://pith.science/pith/A2QH4JN5CKTN2L2K3V4ZKPLURF/action/citation_signature","submit_replication":"https://pith.science/pith/A2QH4JN5CKTN2L2K3V4ZKPLURF/action/replication_record"}},"created_at":"2026-07-05T05:37:13.520643+00:00","updated_at":"2026-07-05T05:37:13.520643+00:00"}