{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:2RWMH544MSMPQ4XGZCTFKK3KZD","short_pith_number":"pith:2RWMH544","schema_version":"1.0","canonical_sha256":"d46cc3f79c6498f872e6c8a6552b6ac8d8a7aae58510ca9d59503a34fab4ac00","source":{"kind":"arxiv","id":"2502.02260","version":2},"attestation_state":"computed","paper":{"title":"Position: Adversarial ML for LLMs Is Not Making Any Progress","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.LG","authors_text":"Florian Tram\\`er, Javier Rando, Jie Zhang, Nicholas Carlini","submitted_at":"2025-02-04T12:17:08Z","abstract_excerpt":"In the past decade, considerable research effort has been devoted to securing machine learning (ML) models that operate in adversarial settings. Yet, progress has been slow even for simple \"toy\" problems (e.g., robustness to small adversarial perturbations) and is often hindered by non-rigorous evaluations. Today, adversarial ML research has shifted towards studying larger, general-purpose language models. In this position paper, we argue that the situation is now even worse: in the era of LLMs, the field of adversarial ML studies problems that are (1) less clearly defined, (2) harder to solve"},"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":"2502.02260","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-02-04T12:17:08Z","cross_cats_sorted":["cs.CR"],"title_canon_sha256":"157cd6184bf56185e93c263f09edf7e8d6fc55a11cfc8e5a4d3fd318f0d41eae","abstract_canon_sha256":"3c0788511a085065172675add35185c48a04fa5c6b86d2211ed7b94d29ba3998"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-03T02:05:40.543970Z","signature_b64":"1qhOkRyMlXSIa0S35wQf3FRqVeIIsx74/V+aKcI467EP6sSkR7bcNcuvM+RNbtHCQkat6eVi4orNZqMBgusBBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d46cc3f79c6498f872e6c8a6552b6ac8d8a7aae58510ca9d59503a34fab4ac00","last_reissued_at":"2026-06-03T02:05:40.543485Z","signature_status":"signed_v1","first_computed_at":"2026-06-03T02:05:40.543485Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Position: Adversarial ML for LLMs Is Not Making Any Progress","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.LG","authors_text":"Florian Tram\\`er, Javier Rando, Jie Zhang, Nicholas Carlini","submitted_at":"2025-02-04T12:17:08Z","abstract_excerpt":"In the past decade, considerable research effort has been devoted to securing machine learning (ML) models that operate in adversarial settings. Yet, progress has been slow even for simple \"toy\" problems (e.g., robustness to small adversarial perturbations) and is often hindered by non-rigorous evaluations. Today, adversarial ML research has shifted towards studying larger, general-purpose language models. In this position paper, we argue that the situation is now even worse: in the era of LLMs, the field of adversarial ML studies problems that are (1) less clearly defined, (2) harder to solve"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.02260","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/2502.02260/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":"2502.02260","created_at":"2026-06-03T02:05:40.543542+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.02260v2","created_at":"2026-06-03T02:05:40.543542+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.02260","created_at":"2026-06-03T02:05:40.543542+00:00"},{"alias_kind":"pith_short_12","alias_value":"2RWMH544MSMP","created_at":"2026-06-03T02:05:40.543542+00:00"},{"alias_kind":"pith_short_16","alias_value":"2RWMH544MSMPQ4XG","created_at":"2026-06-03T02:05:40.543542+00:00"},{"alias_kind":"pith_short_8","alias_value":"2RWMH544","created_at":"2026-06-03T02:05:40.543542+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2607.07918","citing_title":"Efficient Safety Alignment of Language Models via Latent Personality Traits","ref_index":36,"is_internal_anchor":true},{"citing_arxiv_id":"2506.06414","citing_title":"Benchmarking Misuse Mitigation Against Covert Adversaries","ref_index":69,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/2RWMH544MSMPQ4XGZCTFKK3KZD","json":"https://pith.science/pith/2RWMH544MSMPQ4XGZCTFKK3KZD.json","graph_json":"https://pith.science/api/pith-number/2RWMH544MSMPQ4XGZCTFKK3KZD/graph.json","events_json":"https://pith.science/api/pith-number/2RWMH544MSMPQ4XGZCTFKK3KZD/events.json","paper":"https://pith.science/paper/2RWMH544"},"agent_actions":{"view_html":"https://pith.science/pith/2RWMH544MSMPQ4XGZCTFKK3KZD","download_json":"https://pith.science/pith/2RWMH544MSMPQ4XGZCTFKK3KZD.json","view_paper":"https://pith.science/paper/2RWMH544","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.02260&json=true","fetch_graph":"https://pith.science/api/pith-number/2RWMH544MSMPQ4XGZCTFKK3KZD/graph.json","fetch_events":"https://pith.science/api/pith-number/2RWMH544MSMPQ4XGZCTFKK3KZD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2RWMH544MSMPQ4XGZCTFKK3KZD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2RWMH544MSMPQ4XGZCTFKK3KZD/action/storage_attestation","attest_author":"https://pith.science/pith/2RWMH544MSMPQ4XGZCTFKK3KZD/action/author_attestation","sign_citation":"https://pith.science/pith/2RWMH544MSMPQ4XGZCTFKK3KZD/action/citation_signature","submit_replication":"https://pith.science/pith/2RWMH544MSMPQ4XGZCTFKK3KZD/action/replication_record"}},"created_at":"2026-06-03T02:05:40.543542+00:00","updated_at":"2026-06-03T02:05:40.543542+00:00"}