{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:TBLOONOKQIX7L4CGZOPLODJBBL","short_pith_number":"pith:TBLOONOK","canonical_record":{"source":{"id":"2506.01064","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-06-01T16:07:30Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"44a07cd54c3fd8e980874c79b04c0b9d84858e7425819bf4b6f2874bfefbe7c4","abstract_canon_sha256":"dd9a77b10e13d9a846dcadc470523405f2d91561d92cad07132b4be72c7bb7f7"},"schema_version":"1.0"},"canonical_sha256":"9856e735ca822ff5f046cb9eb70d210afc21445e5a235c6b9970ab49ca3d413c","source":{"kind":"arxiv","id":"2506.01064","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.01064","created_at":"2026-07-05T12:11:18Z"},{"alias_kind":"arxiv_version","alias_value":"2506.01064v3","created_at":"2026-07-05T12:11:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.01064","created_at":"2026-07-05T12:11:18Z"},{"alias_kind":"pith_short_12","alias_value":"TBLOONOKQIX7","created_at":"2026-07-05T12:11:18Z"},{"alias_kind":"pith_short_16","alias_value":"TBLOONOKQIX7L4CG","created_at":"2026-07-05T12:11:18Z"},{"alias_kind":"pith_short_8","alias_value":"TBLOONOK","created_at":"2026-07-05T12:11:18Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:TBLOONOKQIX7L4CGZOPLODJBBL","target":"record","payload":{"canonical_record":{"source":{"id":"2506.01064","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-06-01T16:07:30Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"44a07cd54c3fd8e980874c79b04c0b9d84858e7425819bf4b6f2874bfefbe7c4","abstract_canon_sha256":"dd9a77b10e13d9a846dcadc470523405f2d91561d92cad07132b4be72c7bb7f7"},"schema_version":"1.0"},"canonical_sha256":"9856e735ca822ff5f046cb9eb70d210afc21445e5a235c6b9970ab49ca3d413c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:11:18.394992Z","signature_b64":"ndQgO0Vo3uQjso0ENvwb0PYmqp23XzoEIZDHi0/VJQnXiShmVcO7FZe8URQoMemVBfTfOo0gs4lNxE4oKX6GCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9856e735ca822ff5f046cb9eb70d210afc21445e5a235c6b9970ab49ca3d413c","last_reissued_at":"2026-07-05T12:11:18.394411Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:11:18.394411Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2506.01064","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T12:11:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"LYwGPcnZ7YHmZapu51R0zmbQg64fgcLcWQWSKHfjC86QCAsYj9b/lV+s2kV4KyXiSXhi174Fq+B7ryMk2ED0BA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T20:55:13.159084Z"},"content_sha256":"b8e6b858cbc03ad720e53d555353e239a6b48cc0ff99441a7ffe9893588e84ff","schema_version":"1.0","event_id":"sha256:b8e6b858cbc03ad720e53d555353e239a6b48cc0ff99441a7ffe9893588e84ff"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:TBLOONOKQIX7L4CGZOPLODJBBL","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Fighting Fire with Fire (F3): A Training-free and Efficient Visual Adversarial Example Purification Method in LVLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Di Wang, Jiansheng Chen, Ruobing Xie, Xingwu Sun, Yiqing Huang, Yudong Zhang, Yu Wang, Zhanhui Kang","submitted_at":"2025-06-01T16:07:30Z","abstract_excerpt":"Recent advances in large vision-language models (LVLMs) have showcased their remarkable capabilities across a wide range of multimodal vision-language tasks. However, these models remain vulnerable to visual adversarial attacks, which can substantially compromise their performance. In this paper, we introduce F3, a novel adversarial purification framework that employs a counterintuitive ``fighting fire with fire'' strategy: intentionally introducing simple perturbations to adversarial examples to mitigate their harmful effects. Specifically, F3 leverages cross-modal attentions derived from ran"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.01064","kind":"arxiv","version":3},"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/2506.01064/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T12:11:18Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"JlfA0iDgVsMurggdq+p7bGqx4GuJQRuM5nGuo2JUMT8H4cfZaU+5GkBCoo/Acf17MsfdRgM4KGEsAlxjzjaHCw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T20:55:13.160081Z"},"content_sha256":"a9c3fa9daca60c6298f3d8c725448edeb8c223514931737204c2d3f3c9d12ca7","schema_version":"1.0","event_id":"sha256:a9c3fa9daca60c6298f3d8c725448edeb8c223514931737204c2d3f3c9d12ca7"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/TBLOONOKQIX7L4CGZOPLODJBBL/bundle.json","state_url":"https://pith.science/pith/TBLOONOKQIX7L4CGZOPLODJBBL/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/TBLOONOKQIX7L4CGZOPLODJBBL/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-07T20:55:13Z","links":{"resolver":"https://pith.science/pith/TBLOONOKQIX7L4CGZOPLODJBBL","bundle":"https://pith.science/pith/TBLOONOKQIX7L4CGZOPLODJBBL/bundle.json","state":"https://pith.science/pith/TBLOONOKQIX7L4CGZOPLODJBBL/state.json","well_known_bundle":"https://pith.science/.well-known/pith/TBLOONOKQIX7L4CGZOPLODJBBL/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:TBLOONOKQIX7L4CGZOPLODJBBL","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":"dd9a77b10e13d9a846dcadc470523405f2d91561d92cad07132b4be72c7bb7f7","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-06-01T16:07:30Z","title_canon_sha256":"44a07cd54c3fd8e980874c79b04c0b9d84858e7425819bf4b6f2874bfefbe7c4"},"schema_version":"1.0","source":{"id":"2506.01064","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2506.01064","created_at":"2026-07-05T12:11:18Z"},{"alias_kind":"arxiv_version","alias_value":"2506.01064v3","created_at":"2026-07-05T12:11:18Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.01064","created_at":"2026-07-05T12:11:18Z"},{"alias_kind":"pith_short_12","alias_value":"TBLOONOKQIX7","created_at":"2026-07-05T12:11:18Z"},{"alias_kind":"pith_short_16","alias_value":"TBLOONOKQIX7L4CG","created_at":"2026-07-05T12:11:18Z"},{"alias_kind":"pith_short_8","alias_value":"TBLOONOK","created_at":"2026-07-05T12:11:18Z"}],"graph_snapshots":[{"event_id":"sha256:a9c3fa9daca60c6298f3d8c725448edeb8c223514931737204c2d3f3c9d12ca7","target":"graph","created_at":"2026-07-05T12:11:18Z","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/2506.01064/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recent advances in large vision-language models (LVLMs) have showcased their remarkable capabilities across a wide range of multimodal vision-language tasks. However, these models remain vulnerable to visual adversarial attacks, which can substantially compromise their performance. In this paper, we introduce F3, a novel adversarial purification framework that employs a counterintuitive ``fighting fire with fire'' strategy: intentionally introducing simple perturbations to adversarial examples to mitigate their harmful effects. Specifically, F3 leverages cross-modal attentions derived from ran","authors_text":"Di Wang, Jiansheng Chen, Ruobing Xie, Xingwu Sun, Yiqing Huang, Yudong Zhang, Yu Wang, Zhanhui Kang","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-06-01T16:07:30Z","title":"Fighting Fire with Fire (F3): A Training-free and Efficient Visual Adversarial Example Purification Method in LVLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.01064","kind":"arxiv","version":3},"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:b8e6b858cbc03ad720e53d555353e239a6b48cc0ff99441a7ffe9893588e84ff","target":"record","created_at":"2026-07-05T12:11:18Z","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":"dd9a77b10e13d9a846dcadc470523405f2d91561d92cad07132b4be72c7bb7f7","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-06-01T16:07:30Z","title_canon_sha256":"44a07cd54c3fd8e980874c79b04c0b9d84858e7425819bf4b6f2874bfefbe7c4"},"schema_version":"1.0","source":{"id":"2506.01064","kind":"arxiv","version":3}},"canonical_sha256":"9856e735ca822ff5f046cb9eb70d210afc21445e5a235c6b9970ab49ca3d413c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9856e735ca822ff5f046cb9eb70d210afc21445e5a235c6b9970ab49ca3d413c","first_computed_at":"2026-07-05T12:11:18.394411Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T12:11:18.394411Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"ndQgO0Vo3uQjso0ENvwb0PYmqp23XzoEIZDHi0/VJQnXiShmVcO7FZe8URQoMemVBfTfOo0gs4lNxE4oKX6GCg==","signature_status":"signed_v1","signed_at":"2026-07-05T12:11:18.394992Z","signed_message":"canonical_sha256_bytes"},"source_id":"2506.01064","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b8e6b858cbc03ad720e53d555353e239a6b48cc0ff99441a7ffe9893588e84ff","sha256:a9c3fa9daca60c6298f3d8c725448edeb8c223514931737204c2d3f3c9d12ca7"],"state_sha256":"dbadcbc93a9f926b094cd892d079df52b7a1a17b3148ffc8f368ce9853794bdb"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FW4/ggAZvcNSkSdn7hObcek0Bw0SsV8KPrZv5ubaix5hfDzTqsN1meYpBHBlSJm5ja+/EdjXameuTD1DXFwrCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T20:55:13.167221Z","bundle_sha256":"051c5e16b1114cf42a07973bdaef795dca11ddc47e440aec80aac81e5b500f8d"}}