{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:4M43DOUNAPVLL6BNIMEELRB2XD","short_pith_number":"pith:4M43DOUN","schema_version":"1.0","canonical_sha256":"e339b1ba8d03eab5f82d430845c43ab8eec0c4ed9da8a3f2a38019ec3f89dbed","source":{"kind":"arxiv","id":"2106.06946","version":3},"attestation_state":"computed","paper":{"title":"Boosting Randomized Smoothing with Variance Reduced Classifiers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.LG","authors_text":"Marc Fischer, Mark Niklas M\\\"uller, Martin Vechev, Mikl\\'os Z. Horv\\'ath","submitted_at":"2021-06-13T08:40:27Z","abstract_excerpt":"Randomized Smoothing (RS) is a promising method for obtaining robustness certificates by evaluating a base model under noise. In this work, we: (i) theoretically motivate why ensembles are a particularly suitable choice as base models for RS, and (ii) empirically confirm this choice, obtaining state-of-the-art results in multiple settings. The key insight of our work is that the reduced variance of ensembles over the perturbations introduced in RS leads to significantly more consistent classifications for a given input. This, in turn, leads to substantially increased certifiable radii for samp"},"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":"2106.06946","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2021-06-13T08:40:27Z","cross_cats_sorted":["cs.AI","cs.CV"],"title_canon_sha256":"c6a590d40494e53993bfcf17879598722cbfb8c8bab0adbf8ddc8558db4a934e","abstract_canon_sha256":"1f6ac62b338b296e9e4b36eb867ee60d9f46f2706ae563d2b7831626cb81f5a1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:09:56.863498Z","signature_b64":"nvbmc6G4ZgU7DEHbO8stEEWhOC3eLQrGqW/w7vyBj70UDosKAS6iswQv9AH8M5R3pxKjg+OIBoT1miJSNMG4CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e339b1ba8d03eab5f82d430845c43ab8eec0c4ed9da8a3f2a38019ec3f89dbed","last_reissued_at":"2026-07-05T04:09:56.863052Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:09:56.863052Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Boosting Randomized Smoothing with Variance Reduced Classifiers","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CV"],"primary_cat":"cs.LG","authors_text":"Marc Fischer, Mark Niklas M\\\"uller, Martin Vechev, Mikl\\'os Z. Horv\\'ath","submitted_at":"2021-06-13T08:40:27Z","abstract_excerpt":"Randomized Smoothing (RS) is a promising method for obtaining robustness certificates by evaluating a base model under noise. In this work, we: (i) theoretically motivate why ensembles are a particularly suitable choice as base models for RS, and (ii) empirically confirm this choice, obtaining state-of-the-art results in multiple settings. The key insight of our work is that the reduced variance of ensembles over the perturbations introduced in RS leads to significantly more consistent classifications for a given input. This, in turn, leads to substantially increased certifiable radii for samp"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2106.06946","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/2106.06946/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":"2106.06946","created_at":"2026-07-05T04:09:56.863111+00:00"},{"alias_kind":"arxiv_version","alias_value":"2106.06946v3","created_at":"2026-07-05T04:09:56.863111+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2106.06946","created_at":"2026-07-05T04:09:56.863111+00:00"},{"alias_kind":"pith_short_12","alias_value":"4M43DOUNAPVL","created_at":"2026-07-05T04:09:56.863111+00:00"},{"alias_kind":"pith_short_16","alias_value":"4M43DOUNAPVLL6BN","created_at":"2026-07-05T04:09:56.863111+00:00"},{"alias_kind":"pith_short_8","alias_value":"4M43DOUN","created_at":"2026-07-05T04:09:56.863111+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.07690","citing_title":"Fortifying Time Series: DTW-Certified Robust Anomaly Detection","ref_index":27,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/4M43DOUNAPVLL6BNIMEELRB2XD","json":"https://pith.science/pith/4M43DOUNAPVLL6BNIMEELRB2XD.json","graph_json":"https://pith.science/api/pith-number/4M43DOUNAPVLL6BNIMEELRB2XD/graph.json","events_json":"https://pith.science/api/pith-number/4M43DOUNAPVLL6BNIMEELRB2XD/events.json","paper":"https://pith.science/paper/4M43DOUN"},"agent_actions":{"view_html":"https://pith.science/pith/4M43DOUNAPVLL6BNIMEELRB2XD","download_json":"https://pith.science/pith/4M43DOUNAPVLL6BNIMEELRB2XD.json","view_paper":"https://pith.science/paper/4M43DOUN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2106.06946&json=true","fetch_graph":"https://pith.science/api/pith-number/4M43DOUNAPVLL6BNIMEELRB2XD/graph.json","fetch_events":"https://pith.science/api/pith-number/4M43DOUNAPVLL6BNIMEELRB2XD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4M43DOUNAPVLL6BNIMEELRB2XD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4M43DOUNAPVLL6BNIMEELRB2XD/action/storage_attestation","attest_author":"https://pith.science/pith/4M43DOUNAPVLL6BNIMEELRB2XD/action/author_attestation","sign_citation":"https://pith.science/pith/4M43DOUNAPVLL6BNIMEELRB2XD/action/citation_signature","submit_replication":"https://pith.science/pith/4M43DOUNAPVLL6BNIMEELRB2XD/action/replication_record"}},"created_at":"2026-07-05T04:09:56.863111+00:00","updated_at":"2026-07-05T04:09:56.863111+00:00"}