{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:GJSDYKC365RUHSOBFQBSXXOIQF","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"6cea70636da2f5080e3e9f6be231c60c6ce82a77a6e2f5fb4d2e2596415f56f4","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2025-01-16T10:20:48Z","title_canon_sha256":"b8ddd1568ed45ca6802dd9e3ea02424fbbb217594b6e593bec4003d0de1bfa59"},"schema_version":"1.0","source":{"id":"2501.09446","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.09446","created_at":"2026-07-05T10:45:38Z"},{"alias_kind":"arxiv_version","alias_value":"2501.09446v2","created_at":"2026-07-05T10:45:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.09446","created_at":"2026-07-05T10:45:38Z"},{"alias_kind":"pith_short_12","alias_value":"GJSDYKC365RU","created_at":"2026-07-05T10:45:38Z"},{"alias_kind":"pith_short_16","alias_value":"GJSDYKC365RUHSOB","created_at":"2026-07-05T10:45:38Z"},{"alias_kind":"pith_short_8","alias_value":"GJSDYKC3","created_at":"2026-07-05T10:45:38Z"}],"graph_snapshots":[{"event_id":"sha256:ace73f603834ef115ade413dc9339cabcd62c6fe8f0aae1e72dc6e8c1c318859","target":"graph","created_at":"2026-07-05T10:45:38Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2501.09446/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"This paper investigates the robustness of vision-language models against adversarial visual perturbations and introduces a novel ``double visual defense\" to enhance this robustness. Unlike previous approaches that resort to lightweight adversarial fine-tuning of a pre-trained CLIP model, we perform large-scale adversarial vision-language pre-training from scratch using web-scale data. We then strengthen the defense by incorporating adversarial visual instruction tuning. The resulting models from each stage, $\\Delta$CLIP and $\\Delta^2$LLaVA, show substantially enhanced zero-shot robustness and ","authors_text":"Bhavya Kailkhura, Brian Bartoldson, Cihang Xie, Zeyu Wang","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2025-01-16T10:20:48Z","title":"Double Visual Defense: Adversarial Pre-training and Instruction Tuning for Improving Vision-Language Model Robustness"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.09446","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:ad0288f89fa0f90da0b1c8f30e07ace6de7aa690c8153049ab6765d6c105575f","target":"record","created_at":"2026-07-05T10:45:38Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"6cea70636da2f5080e3e9f6be231c60c6ce82a77a6e2f5fb4d2e2596415f56f4","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2025-01-16T10:20:48Z","title_canon_sha256":"b8ddd1568ed45ca6802dd9e3ea02424fbbb217594b6e593bec4003d0de1bfa59"},"schema_version":"1.0","source":{"id":"2501.09446","kind":"arxiv","version":2}},"canonical_sha256":"32643c285bf76343c9c12c032bddc88170b08013cd9354e9a6136840cefb4fea","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"32643c285bf76343c9c12c032bddc88170b08013cd9354e9a6136840cefb4fea","first_computed_at":"2026-07-05T10:45:38.899181Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:45:38.899181Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"mH0B/gw44ZqQB3mgAME8+LloxkbnBEhO7owMr2okHRJRe+elnDFwTFQV978y7/B1T6yL0XOBmsyNBgC7A6L9Bw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:45:38.899674Z","signed_message":"canonical_sha256_bytes"},"source_id":"2501.09446","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:ad0288f89fa0f90da0b1c8f30e07ace6de7aa690c8153049ab6765d6c105575f","sha256:ace73f603834ef115ade413dc9339cabcd62c6fe8f0aae1e72dc6e8c1c318859"],"state_sha256":"2c6b16a26055811bac45bc75cc7ffccc0d5c88a1606f0112053c6e2f9e163b66"}