{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:X3PHAEEZGKYOM2HS4R3UGRZGIS","short_pith_number":"pith:X3PHAEEZ","schema_version":"1.0","canonical_sha256":"bede70109932b0e668f2e47743472644a2abecd4d3146e53354ae3c852171950","source":{"kind":"arxiv","id":"2410.21802","version":2},"attestation_state":"computed","paper":{"title":"Text-Guided Attention is All You Need for Zero-Shot Robustness in Vision-Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Changsheng Xu, Haiyang Zhang, Lu Yu","submitted_at":"2024-10-29T07:15:09Z","abstract_excerpt":"Due to the impressive zero-shot capabilities, pre-trained vision-language models (e.g. CLIP), have attracted widespread attention and adoption across various domains. Nonetheless, CLIP has been observed to be susceptible to adversarial examples. Through experimental analysis, we have observed a phenomenon wherein adversarial perturbations induce shifts in text-guided attention. Building upon this observation, we propose a simple yet effective strategy: Text-Guided Attention for Zero-Shot Robustness (TGA-ZSR). This framework incorporates two components: the Attention Refinement module and the A"},"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":"2410.21802","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-10-29T07:15:09Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"0b608cb1c98107f09753df470ab5f3a9b91093cc9fe11f44f41af1838b99ad23","abstract_canon_sha256":"4cd1f6128939422f37061f2b587864dd5dedcbcec76a88a314bf580ccaa1b836"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:28:14.652411Z","signature_b64":"C0Q7PxMGIjrShVPocwWEL47FW9Pb1of+rNk7jWhfkb/awdWPqthGsKHCoZSmBzSZPjOFKNzqd2XyKt0sn0iUCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bede70109932b0e668f2e47743472644a2abecd4d3146e53354ae3c852171950","last_reissued_at":"2026-07-05T09:28:14.651962Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:28:14.651962Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Text-Guided Attention is All You Need for Zero-Shot Robustness in Vision-Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Changsheng Xu, Haiyang Zhang, Lu Yu","submitted_at":"2024-10-29T07:15:09Z","abstract_excerpt":"Due to the impressive zero-shot capabilities, pre-trained vision-language models (e.g. CLIP), have attracted widespread attention and adoption across various domains. Nonetheless, CLIP has been observed to be susceptible to adversarial examples. Through experimental analysis, we have observed a phenomenon wherein adversarial perturbations induce shifts in text-guided attention. Building upon this observation, we propose a simple yet effective strategy: Text-Guided Attention for Zero-Shot Robustness (TGA-ZSR). This framework incorporates two components: the Attention Refinement module and the A"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.21802","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/2410.21802/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":"2410.21802","created_at":"2026-07-05T09:28:14.652015+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.21802v2","created_at":"2026-07-05T09:28:14.652015+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.21802","created_at":"2026-07-05T09:28:14.652015+00:00"},{"alias_kind":"pith_short_12","alias_value":"X3PHAEEZGKYO","created_at":"2026-07-05T09:28:14.652015+00:00"},{"alias_kind":"pith_short_16","alias_value":"X3PHAEEZGKYOM2HS","created_at":"2026-07-05T09:28:14.652015+00:00"},{"alias_kind":"pith_short_8","alias_value":"X3PHAEEZ","created_at":"2026-07-05T09:28:14.652015+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/X3PHAEEZGKYOM2HS4R3UGRZGIS","json":"https://pith.science/pith/X3PHAEEZGKYOM2HS4R3UGRZGIS.json","graph_json":"https://pith.science/api/pith-number/X3PHAEEZGKYOM2HS4R3UGRZGIS/graph.json","events_json":"https://pith.science/api/pith-number/X3PHAEEZGKYOM2HS4R3UGRZGIS/events.json","paper":"https://pith.science/paper/X3PHAEEZ"},"agent_actions":{"view_html":"https://pith.science/pith/X3PHAEEZGKYOM2HS4R3UGRZGIS","download_json":"https://pith.science/pith/X3PHAEEZGKYOM2HS4R3UGRZGIS.json","view_paper":"https://pith.science/paper/X3PHAEEZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.21802&json=true","fetch_graph":"https://pith.science/api/pith-number/X3PHAEEZGKYOM2HS4R3UGRZGIS/graph.json","fetch_events":"https://pith.science/api/pith-number/X3PHAEEZGKYOM2HS4R3UGRZGIS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/X3PHAEEZGKYOM2HS4R3UGRZGIS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/X3PHAEEZGKYOM2HS4R3UGRZGIS/action/storage_attestation","attest_author":"https://pith.science/pith/X3PHAEEZGKYOM2HS4R3UGRZGIS/action/author_attestation","sign_citation":"https://pith.science/pith/X3PHAEEZGKYOM2HS4R3UGRZGIS/action/citation_signature","submit_replication":"https://pith.science/pith/X3PHAEEZGKYOM2HS4R3UGRZGIS/action/replication_record"}},"created_at":"2026-07-05T09:28:14.652015+00:00","updated_at":"2026-07-05T09:28:14.652015+00:00"}