{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:OXXEGZ2KH5LA6O4TSKDRSSCING","short_pith_number":"pith:OXXEGZ2K","schema_version":"1.0","canonical_sha256":"75ee43674a3f560f3b939287194848698935b863f28472c72f9a01099dac6b60","source":{"kind":"arxiv","id":"1812.02637","version":4},"attestation_state":"computed","paper":{"title":"MMA Training: Direct Input Space Margin Maximization through Adversarial Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NE","stat.ML"],"primary_cat":"cs.LG","authors_text":"Gavin Weiguang Ding, Kry Yik Chau Lui, Ruitong Huang, Yash Sharma","submitted_at":"2018-12-06T16:15:52Z","abstract_excerpt":"We study adversarial robustness of neural networks from a margin maximization perspective, where margins are defined as the distances from inputs to a classifier's decision boundary. Our study shows that maximizing margins can be achieved by minimizing the adversarial loss on the decision boundary at the \"shortest successful perturbation\", demonstrating a close connection between adversarial losses and the margins. We propose Max-Margin Adversarial (MMA) training to directly maximize the margins to achieve adversarial robustness. Instead of adversarial training with a fixed $\\epsilon$, MMA off"},"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":"1812.02637","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-12-06T16:15:52Z","cross_cats_sorted":["cs.NE","stat.ML"],"title_canon_sha256":"886f6c785d1e7bfe22ea9ab8395ab76405aa7005ec7a7aca69fb2b265c99e142","abstract_canon_sha256":"dab979b8e66d325913571c70e7d6849d380e7a9abf812771ca9a8d0c2d5b56a8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:45:50.153748Z","signature_b64":"ClTfkMn+Ydn1rALjc/35EPp3ZiiBCTG7iO9z+8gcpUxIejDQ94KRakq4m+zVsu58BPXF4y6t9UVXjcreSCxfBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"75ee43674a3f560f3b939287194848698935b863f28472c72f9a01099dac6b60","last_reissued_at":"2026-07-05T00:45:50.153297Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:45:50.153297Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MMA Training: Direct Input Space Margin Maximization through Adversarial Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.NE","stat.ML"],"primary_cat":"cs.LG","authors_text":"Gavin Weiguang Ding, Kry Yik Chau Lui, Ruitong Huang, Yash Sharma","submitted_at":"2018-12-06T16:15:52Z","abstract_excerpt":"We study adversarial robustness of neural networks from a margin maximization perspective, where margins are defined as the distances from inputs to a classifier's decision boundary. Our study shows that maximizing margins can be achieved by minimizing the adversarial loss on the decision boundary at the \"shortest successful perturbation\", demonstrating a close connection between adversarial losses and the margins. We propose Max-Margin Adversarial (MMA) training to directly maximize the margins to achieve adversarial robustness. Instead of adversarial training with a fixed $\\epsilon$, MMA off"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1812.02637","kind":"arxiv","version":4},"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/1812.02637/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":"1812.02637","created_at":"2026-07-05T00:45:50.153367+00:00"},{"alias_kind":"arxiv_version","alias_value":"1812.02637v4","created_at":"2026-07-05T00:45:50.153367+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1812.02637","created_at":"2026-07-05T00:45:50.153367+00:00"},{"alias_kind":"pith_short_12","alias_value":"OXXEGZ2KH5LA","created_at":"2026-07-05T00:45:50.153367+00:00"},{"alias_kind":"pith_short_16","alias_value":"OXXEGZ2KH5LA6O4T","created_at":"2026-07-05T00:45:50.153367+00:00"},{"alias_kind":"pith_short_8","alias_value":"OXXEGZ2K","created_at":"2026-07-05T00:45:50.153367+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.00738","citing_title":"SORA: Free Second-Order Attacks in Fast Adversarial Training","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15416","citing_title":"Margin-Adaptive Confidence Ranking for Reliable LLM Judgement","ref_index":54,"is_internal_anchor":false},{"citing_arxiv_id":"2604.20619","citing_title":"Stochastic Krylov Dynamics: Revisiting Operator Growth in Open Quantum Systems","ref_index":11,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/OXXEGZ2KH5LA6O4TSKDRSSCING","json":"https://pith.science/pith/OXXEGZ2KH5LA6O4TSKDRSSCING.json","graph_json":"https://pith.science/api/pith-number/OXXEGZ2KH5LA6O4TSKDRSSCING/graph.json","events_json":"https://pith.science/api/pith-number/OXXEGZ2KH5LA6O4TSKDRSSCING/events.json","paper":"https://pith.science/paper/OXXEGZ2K"},"agent_actions":{"view_html":"https://pith.science/pith/OXXEGZ2KH5LA6O4TSKDRSSCING","download_json":"https://pith.science/pith/OXXEGZ2KH5LA6O4TSKDRSSCING.json","view_paper":"https://pith.science/paper/OXXEGZ2K","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1812.02637&json=true","fetch_graph":"https://pith.science/api/pith-number/OXXEGZ2KH5LA6O4TSKDRSSCING/graph.json","fetch_events":"https://pith.science/api/pith-number/OXXEGZ2KH5LA6O4TSKDRSSCING/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OXXEGZ2KH5LA6O4TSKDRSSCING/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OXXEGZ2KH5LA6O4TSKDRSSCING/action/storage_attestation","attest_author":"https://pith.science/pith/OXXEGZ2KH5LA6O4TSKDRSSCING/action/author_attestation","sign_citation":"https://pith.science/pith/OXXEGZ2KH5LA6O4TSKDRSSCING/action/citation_signature","submit_replication":"https://pith.science/pith/OXXEGZ2KH5LA6O4TSKDRSSCING/action/replication_record"}},"created_at":"2026-07-05T00:45:50.153367+00:00","updated_at":"2026-07-05T00:45:50.153367+00:00"}