{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:573KORMIPISHEVU4QGO6LAFGV4","short_pith_number":"pith:573KORMI","schema_version":"1.0","canonical_sha256":"eff6a745887a2472569c819de580a6af140a722cf2395d957d841adfaff114ed","source":{"kind":"arxiv","id":"2509.09534","version":1},"attestation_state":"computed","paper":{"title":"ProDiGy: Proximity- and Dissimilarity-Based Byzantine-Robust Federated Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.DC"],"primary_cat":"cs.LG","authors_text":"Luis Ma{\\ss}ny, Rawad Bitar, Sena Ergisi","submitted_at":"2025-09-11T15:25:59Z","abstract_excerpt":"Federated Learning (FL) emerged as a widely studied paradigm for distributed learning. Despite its many advantages, FL remains vulnerable to adversarial attacks, especially under data heterogeneity. We propose a new Byzantine-robust FL algorithm called ProDiGy. The key novelty lies in evaluating the client gradients using a joint dual scoring system based on the gradients' proximity and dissimilarity. We demonstrate through extensive numerical experiments that ProDiGy outperforms existing defenses in various scenarios. In particular, when the clients' data do not follow an IID distribution, wh"},"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":"2509.09534","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-09-11T15:25:59Z","cross_cats_sorted":["cs.DC"],"title_canon_sha256":"c85908b80991c7829818281fb3ca7581b05e73984f88dd6245a59c3a3d131ebe","abstract_canon_sha256":"f9ad42d9a3db1077abebb9ad43e152429ec43a5901c94db7059eac191fa9a5c2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:09:37.641192Z","signature_b64":"6sI//KE4aIle3Qv2AGQuq/jT89zRV+xQwI5UMMg2cylkubV9vzcKm2Sbh4EpYthXHWt4cLp7deXOrnMfgOR6Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"eff6a745887a2472569c819de580a6af140a722cf2395d957d841adfaff114ed","last_reissued_at":"2026-07-05T12:09:37.640611Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:09:37.640611Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ProDiGy: Proximity- and Dissimilarity-Based Byzantine-Robust Federated Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.DC"],"primary_cat":"cs.LG","authors_text":"Luis Ma{\\ss}ny, Rawad Bitar, Sena Ergisi","submitted_at":"2025-09-11T15:25:59Z","abstract_excerpt":"Federated Learning (FL) emerged as a widely studied paradigm for distributed learning. Despite its many advantages, FL remains vulnerable to adversarial attacks, especially under data heterogeneity. We propose a new Byzantine-robust FL algorithm called ProDiGy. The key novelty lies in evaluating the client gradients using a joint dual scoring system based on the gradients' proximity and dissimilarity. We demonstrate through extensive numerical experiments that ProDiGy outperforms existing defenses in various scenarios. In particular, when the clients' data do not follow an IID distribution, wh"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.09534","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/2509.09534/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":"2509.09534","created_at":"2026-07-05T12:09:37.640678+00:00"},{"alias_kind":"arxiv_version","alias_value":"2509.09534v1","created_at":"2026-07-05T12:09:37.640678+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.09534","created_at":"2026-07-05T12:09:37.640678+00:00"},{"alias_kind":"pith_short_12","alias_value":"573KORMIPISH","created_at":"2026-07-05T12:09:37.640678+00:00"},{"alias_kind":"pith_short_16","alias_value":"573KORMIPISHEVU4","created_at":"2026-07-05T12:09:37.640678+00:00"},{"alias_kind":"pith_short_8","alias_value":"573KORMI","created_at":"2026-07-05T12:09:37.640678+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/573KORMIPISHEVU4QGO6LAFGV4","json":"https://pith.science/pith/573KORMIPISHEVU4QGO6LAFGV4.json","graph_json":"https://pith.science/api/pith-number/573KORMIPISHEVU4QGO6LAFGV4/graph.json","events_json":"https://pith.science/api/pith-number/573KORMIPISHEVU4QGO6LAFGV4/events.json","paper":"https://pith.science/paper/573KORMI"},"agent_actions":{"view_html":"https://pith.science/pith/573KORMIPISHEVU4QGO6LAFGV4","download_json":"https://pith.science/pith/573KORMIPISHEVU4QGO6LAFGV4.json","view_paper":"https://pith.science/paper/573KORMI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2509.09534&json=true","fetch_graph":"https://pith.science/api/pith-number/573KORMIPISHEVU4QGO6LAFGV4/graph.json","fetch_events":"https://pith.science/api/pith-number/573KORMIPISHEVU4QGO6LAFGV4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/573KORMIPISHEVU4QGO6LAFGV4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/573KORMIPISHEVU4QGO6LAFGV4/action/storage_attestation","attest_author":"https://pith.science/pith/573KORMIPISHEVU4QGO6LAFGV4/action/author_attestation","sign_citation":"https://pith.science/pith/573KORMIPISHEVU4QGO6LAFGV4/action/citation_signature","submit_replication":"https://pith.science/pith/573KORMIPISHEVU4QGO6LAFGV4/action/replication_record"}},"created_at":"2026-07-05T12:09:37.640678+00:00","updated_at":"2026-07-05T12:09:37.640678+00:00"}