{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:GMIM3M5A5C7XVTS4LPCP32SA3I","short_pith_number":"pith:GMIM3M5A","schema_version":"1.0","canonical_sha256":"3310cdb3a0e8bf7ace5c5bc4fdea40da31ecef76a05df58e81e82962015ca801","source":{"kind":"arxiv","id":"2106.07904","version":2},"attestation_state":"computed","paper":{"title":"Probabilistic Margins for Instance Reweighting in Adversarial Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Bo Han, Chen Gong, Feng Liu, Gang Niu, Masashi Sugiyama, Mingyuan Zhou, Qizhou Wang, Tongliang Liu","submitted_at":"2021-06-15T06:37:55Z","abstract_excerpt":"Reweighting adversarial data during training has been recently shown to improve adversarial robustness, where data closer to the current decision boundaries are regarded as more critical and given larger weights. However, existing methods measuring the closeness are not very reliable: they are discrete and can take only a few values, and they are path-dependent, i.e., they may change given the same start and end points with different attack paths. In this paper, we propose three types of probabilistic margin (PM), which are continuous and path-independent, for measuring the aforementioned clos"},"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":"2106.07904","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-06-15T06:37:55Z","cross_cats_sorted":[],"title_canon_sha256":"0e0fe353336f5036c5c38db244c366f3d6da52cd31e2a98fb4eb9c83c9821d57","abstract_canon_sha256":"c134919a48b8760b4c55bf99d0fb83333ab1ac3eeb2a18b45f6be6fbcd06c241"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:05:58.872468Z","signature_b64":"rNc8Wwv3Z4F1HVubsQv3Crmza1+WnaZwpnHqtnearcMJvWEUW97bM6jQ6cRJvSG7p4Tf7LDMYz6ebDjvniZgDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3310cdb3a0e8bf7ace5c5bc4fdea40da31ecef76a05df58e81e82962015ca801","last_reissued_at":"2026-07-05T04:05:58.872015Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:05:58.872015Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Probabilistic Margins for Instance Reweighting in Adversarial Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Bo Han, Chen Gong, Feng Liu, Gang Niu, Masashi Sugiyama, Mingyuan Zhou, Qizhou Wang, Tongliang Liu","submitted_at":"2021-06-15T06:37:55Z","abstract_excerpt":"Reweighting adversarial data during training has been recently shown to improve adversarial robustness, where data closer to the current decision boundaries are regarded as more critical and given larger weights. However, existing methods measuring the closeness are not very reliable: they are discrete and can take only a few values, and they are path-dependent, i.e., they may change given the same start and end points with different attack paths. In this paper, we propose three types of probabilistic margin (PM), which are continuous and path-independent, for measuring the aforementioned clos"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2106.07904","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/2106.07904/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":"2106.07904","created_at":"2026-07-05T04:05:58.872071+00:00"},{"alias_kind":"arxiv_version","alias_value":"2106.07904v2","created_at":"2026-07-05T04:05:58.872071+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2106.07904","created_at":"2026-07-05T04:05:58.872071+00:00"},{"alias_kind":"pith_short_12","alias_value":"GMIM3M5A5C7X","created_at":"2026-07-05T04:05:58.872071+00:00"},{"alias_kind":"pith_short_16","alias_value":"GMIM3M5A5C7XVTS4","created_at":"2026-07-05T04:05:58.872071+00:00"},{"alias_kind":"pith_short_8","alias_value":"GMIM3M5A","created_at":"2026-07-05T04:05:58.872071+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2509.20786","citing_title":"LiLAW: Lightweight Learnable Adaptive Weighting to Learn Sample Difficulty & Improve Noisy Training","ref_index":26,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/GMIM3M5A5C7XVTS4LPCP32SA3I","json":"https://pith.science/pith/GMIM3M5A5C7XVTS4LPCP32SA3I.json","graph_json":"https://pith.science/api/pith-number/GMIM3M5A5C7XVTS4LPCP32SA3I/graph.json","events_json":"https://pith.science/api/pith-number/GMIM3M5A5C7XVTS4LPCP32SA3I/events.json","paper":"https://pith.science/paper/GMIM3M5A"},"agent_actions":{"view_html":"https://pith.science/pith/GMIM3M5A5C7XVTS4LPCP32SA3I","download_json":"https://pith.science/pith/GMIM3M5A5C7XVTS4LPCP32SA3I.json","view_paper":"https://pith.science/paper/GMIM3M5A","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2106.07904&json=true","fetch_graph":"https://pith.science/api/pith-number/GMIM3M5A5C7XVTS4LPCP32SA3I/graph.json","fetch_events":"https://pith.science/api/pith-number/GMIM3M5A5C7XVTS4LPCP32SA3I/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/GMIM3M5A5C7XVTS4LPCP32SA3I/action/timestamp_anchor","attest_storage":"https://pith.science/pith/GMIM3M5A5C7XVTS4LPCP32SA3I/action/storage_attestation","attest_author":"https://pith.science/pith/GMIM3M5A5C7XVTS4LPCP32SA3I/action/author_attestation","sign_citation":"https://pith.science/pith/GMIM3M5A5C7XVTS4LPCP32SA3I/action/citation_signature","submit_replication":"https://pith.science/pith/GMIM3M5A5C7XVTS4LPCP32SA3I/action/replication_record"}},"created_at":"2026-07-05T04:05:58.872071+00:00","updated_at":"2026-07-05T04:05:58.872071+00:00"}