{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:54I73WTDW7E6XW43X473DHAD76","short_pith_number":"pith:54I73WTD","schema_version":"1.0","canonical_sha256":"ef11fdda63b7c9ebdb9bbf3fb19c03ff9b8908393b5005ff3bb66aea57ac35ef","source":{"kind":"arxiv","id":"2302.02432","version":3},"attestation_state":"computed","paper":{"title":"Tighter Information-Theoretic Generalization Bounds from Supersamples","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.IT","cs.LG","math.IT"],"primary_cat":"stat.ML","authors_text":"Yongyi Mao, Ziqiao Wang","submitted_at":"2023-02-05T17:06:27Z","abstract_excerpt":"In this work, we present a variety of novel information-theoretic generalization bounds for learning algorithms, from the supersample setting of Steinke & Zakynthinou (2020)-the setting of the \"conditional mutual information\" framework. Our development exploits projecting the loss pair (obtained from a training instance and a testing instance) down to a single number and correlating loss values with a Rademacher sequence (and its shifted variants). The presented bounds include square-root bounds, fast-rate bounds, including those based on variance and sharpness, and bounds for interpolating al"},"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":"2302.02432","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"stat.ML","submitted_at":"2023-02-05T17:06:27Z","cross_cats_sorted":["cs.IT","cs.LG","math.IT"],"title_canon_sha256":"60c9b4fe9fa0eeec86b2f7f622669e0380cc095bd4d2e520266c1ab775e7495a","abstract_canon_sha256":"a3ebfe3f3865d12cf41fb77c926a3ca0f3855a1f0697ee48b9648edfe8067584"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:20:40.482833Z","signature_b64":"aHXNeq9cbqPIA63FbomkExBSHi4W55w3g1kqk4A7nkHQ8xiwn/FXZDtFEAh2Hrlg14zh0Si+g5/u5YU2CoVfCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ef11fdda63b7c9ebdb9bbf3fb19c03ff9b8908393b5005ff3bb66aea57ac35ef","last_reissued_at":"2026-07-05T06:20:40.482438Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:20:40.482438Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Tighter Information-Theoretic Generalization Bounds from Supersamples","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.IT","cs.LG","math.IT"],"primary_cat":"stat.ML","authors_text":"Yongyi Mao, Ziqiao Wang","submitted_at":"2023-02-05T17:06:27Z","abstract_excerpt":"In this work, we present a variety of novel information-theoretic generalization bounds for learning algorithms, from the supersample setting of Steinke & Zakynthinou (2020)-the setting of the \"conditional mutual information\" framework. Our development exploits projecting the loss pair (obtained from a training instance and a testing instance) down to a single number and correlating loss values with a Rademacher sequence (and its shifted variants). The presented bounds include square-root bounds, fast-rate bounds, including those based on variance and sharpness, and bounds for interpolating al"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2302.02432","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/2302.02432/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":"2302.02432","created_at":"2026-07-05T06:20:40.482494+00:00"},{"alias_kind":"arxiv_version","alias_value":"2302.02432v3","created_at":"2026-07-05T06:20:40.482494+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2302.02432","created_at":"2026-07-05T06:20:40.482494+00:00"},{"alias_kind":"pith_short_12","alias_value":"54I73WTDW7E6","created_at":"2026-07-05T06:20:40.482494+00:00"},{"alias_kind":"pith_short_16","alias_value":"54I73WTDW7E6XW43","created_at":"2026-07-05T06:20:40.482494+00:00"},{"alias_kind":"pith_short_8","alias_value":"54I73WTD","created_at":"2026-07-05T06:20:40.482494+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.07861","citing_title":"Fairness Overfitting in Machine Learning: An Information-Theoretic Perspective","ref_index":70,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/54I73WTDW7E6XW43X473DHAD76","json":"https://pith.science/pith/54I73WTDW7E6XW43X473DHAD76.json","graph_json":"https://pith.science/api/pith-number/54I73WTDW7E6XW43X473DHAD76/graph.json","events_json":"https://pith.science/api/pith-number/54I73WTDW7E6XW43X473DHAD76/events.json","paper":"https://pith.science/paper/54I73WTD"},"agent_actions":{"view_html":"https://pith.science/pith/54I73WTDW7E6XW43X473DHAD76","download_json":"https://pith.science/pith/54I73WTDW7E6XW43X473DHAD76.json","view_paper":"https://pith.science/paper/54I73WTD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2302.02432&json=true","fetch_graph":"https://pith.science/api/pith-number/54I73WTDW7E6XW43X473DHAD76/graph.json","fetch_events":"https://pith.science/api/pith-number/54I73WTDW7E6XW43X473DHAD76/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/54I73WTDW7E6XW43X473DHAD76/action/timestamp_anchor","attest_storage":"https://pith.science/pith/54I73WTDW7E6XW43X473DHAD76/action/storage_attestation","attest_author":"https://pith.science/pith/54I73WTDW7E6XW43X473DHAD76/action/author_attestation","sign_citation":"https://pith.science/pith/54I73WTDW7E6XW43X473DHAD76/action/citation_signature","submit_replication":"https://pith.science/pith/54I73WTDW7E6XW43X473DHAD76/action/replication_record"}},"created_at":"2026-07-05T06:20:40.482494+00:00","updated_at":"2026-07-05T06:20:40.482494+00:00"}