{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:NPNT5JPFJLB5FZQGPIXBGUMKPX","short_pith_number":"pith:NPNT5JPF","schema_version":"1.0","canonical_sha256":"6bdb3ea5e54ac3d2e6067a2e13518a7ddaef05e7a560fef9db5e3a9107a91e78","source":{"kind":"arxiv","id":"2410.04638","version":2},"attestation_state":"computed","paper":{"title":"Provable Weak-to-Strong Generalization via Benign Overfitting","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Anant Sahai, David X. Wu","submitted_at":"2024-10-06T22:10:50Z","abstract_excerpt":"The classic teacher-student model in machine learning posits that a strong teacher supervises a weak student to improve the student's capabilities. We instead consider the inverted situation, where a weak teacher supervises a strong student with imperfect pseudolabels. This paradigm was recently brought forth by Burns et al.'23 and termed \\emph{weak-to-strong generalization}. We theoretically investigate weak-to-strong generalization for binary and multilabel classification in a stylized overparameterized spiked covariance model with Gaussian covariates where the weak teacher's pseudolabels ar"},"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":"2410.04638","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-10-06T22:10:50Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"f62f07e90a64606e145eb63d045d3c3ea41c8ac3cbf77dccebd49f0c787f4e25","abstract_canon_sha256":"7a0a9a8638e1c7a5641f948b3970c5e7b820d8d731a5cf689371c165ff96e4c8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:07:45.150434Z","signature_b64":"sxnn0LBNYjWJMFCJNZN3EPtu84ZkwXV9WyVc6SNRV2dM4wnFrk+aB5e/W9AWVbDQ167brO6+Td1h0C4AY2kdBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6bdb3ea5e54ac3d2e6067a2e13518a7ddaef05e7a560fef9db5e3a9107a91e78","last_reissued_at":"2026-07-05T10:07:45.149984Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:07:45.149984Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Provable Weak-to-Strong Generalization via Benign Overfitting","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Anant Sahai, David X. Wu","submitted_at":"2024-10-06T22:10:50Z","abstract_excerpt":"The classic teacher-student model in machine learning posits that a strong teacher supervises a weak student to improve the student's capabilities. We instead consider the inverted situation, where a weak teacher supervises a strong student with imperfect pseudolabels. This paradigm was recently brought forth by Burns et al.'23 and termed \\emph{weak-to-strong generalization}. We theoretically investigate weak-to-strong generalization for binary and multilabel classification in a stylized overparameterized spiked covariance model with Gaussian covariates where the weak teacher's pseudolabels ar"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.04638","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/2410.04638/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":"2410.04638","created_at":"2026-07-05T10:07:45.150040+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.04638v2","created_at":"2026-07-05T10:07:45.150040+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.04638","created_at":"2026-07-05T10:07:45.150040+00:00"},{"alias_kind":"pith_short_12","alias_value":"NPNT5JPFJLB5","created_at":"2026-07-05T10:07:45.150040+00:00"},{"alias_kind":"pith_short_16","alias_value":"NPNT5JPFJLB5FZQG","created_at":"2026-07-05T10:07:45.150040+00:00"},{"alias_kind":"pith_short_8","alias_value":"NPNT5JPF","created_at":"2026-07-05T10:07:45.150040+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2502.05075","citing_title":"Discrepancies are Virtue: Weak-to-Strong Generalization through Lens of Intrinsic Dimension","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05742","citing_title":"Weak-to-Strong Generalization is Nearly Inevitable (in Linear Models)","ref_index":3,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NPNT5JPFJLB5FZQGPIXBGUMKPX","json":"https://pith.science/pith/NPNT5JPFJLB5FZQGPIXBGUMKPX.json","graph_json":"https://pith.science/api/pith-number/NPNT5JPFJLB5FZQGPIXBGUMKPX/graph.json","events_json":"https://pith.science/api/pith-number/NPNT5JPFJLB5FZQGPIXBGUMKPX/events.json","paper":"https://pith.science/paper/NPNT5JPF"},"agent_actions":{"view_html":"https://pith.science/pith/NPNT5JPFJLB5FZQGPIXBGUMKPX","download_json":"https://pith.science/pith/NPNT5JPFJLB5FZQGPIXBGUMKPX.json","view_paper":"https://pith.science/paper/NPNT5JPF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.04638&json=true","fetch_graph":"https://pith.science/api/pith-number/NPNT5JPFJLB5FZQGPIXBGUMKPX/graph.json","fetch_events":"https://pith.science/api/pith-number/NPNT5JPFJLB5FZQGPIXBGUMKPX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NPNT5JPFJLB5FZQGPIXBGUMKPX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NPNT5JPFJLB5FZQGPIXBGUMKPX/action/storage_attestation","attest_author":"https://pith.science/pith/NPNT5JPFJLB5FZQGPIXBGUMKPX/action/author_attestation","sign_citation":"https://pith.science/pith/NPNT5JPFJLB5FZQGPIXBGUMKPX/action/citation_signature","submit_replication":"https://pith.science/pith/NPNT5JPFJLB5FZQGPIXBGUMKPX/action/replication_record"}},"created_at":"2026-07-05T10:07:45.150040+00:00","updated_at":"2026-07-05T10:07:45.150040+00:00"}