{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:FDZWHDXGOITF4NBPZGAB7UNL2I","short_pith_number":"pith:FDZWHDXG","schema_version":"1.0","canonical_sha256":"28f3638ee672265e342fc9801fd1abd21c4be0b97025fe31907b027c6860fc7a","source":{"kind":"arxiv","id":"2605.19956","version":1},"attestation_state":"computed","paper":{"title":"Towards Fine-Grained Robustness: Attention-Guided Test-Time Prompt Tuning for Vision-Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Jia-Wei Hai, Xiu-Shen Wei, Yijun Wang","submitted_at":"2026-05-19T15:10:06Z","abstract_excerpt":"Vision-Language Models (VLMs), such as CLIP, have achieved significant zero-shot performance on downstream tasks with various fine-tuning adaptation methods. However, recent studies have proven that adversarial attacks can significantly degrade the inference ability of VLMs, posing substantial risks to their practical applications. Prevalent test-time adaptation methods typically rely on multi-view augmentation to implement various fine-tuning strategies, which struggle to identify semantic information and are prone to destroying discriminative regions in fine-grained scenarios. To address the"},"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":"2605.19956","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-05-19T15:10:06Z","cross_cats_sorted":[],"title_canon_sha256":"195fe50c07820ba173f209146d16f78db8d8b68bc9c5256dfd37707ea1b62996","abstract_canon_sha256":"39df0c9b2f03becc21cb33a2fdb2bb720a60d669bdc6d2bf388aaea08630d8d7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T02:05:56.913444Z","signature_b64":"hRGgbkHUOWtFa3jBTa99h8dTPQNxOoTch8Coy4TDeyp93Z7bLvhQ1p6vjsQZYzLyNO2jO1YVudKS0Z5uvZ1dDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"28f3638ee672265e342fc9801fd1abd21c4be0b97025fe31907b027c6860fc7a","last_reissued_at":"2026-05-20T02:05:56.912852Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T02:05:56.912852Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Towards Fine-Grained Robustness: Attention-Guided Test-Time Prompt Tuning for Vision-Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Jia-Wei Hai, Xiu-Shen Wei, Yijun Wang","submitted_at":"2026-05-19T15:10:06Z","abstract_excerpt":"Vision-Language Models (VLMs), such as CLIP, have achieved significant zero-shot performance on downstream tasks with various fine-tuning adaptation methods. However, recent studies have proven that adversarial attacks can significantly degrade the inference ability of VLMs, posing substantial risks to their practical applications. Prevalent test-time adaptation methods typically rely on multi-view augmentation to implement various fine-tuning strategies, which struggle to identify semantic information and are prone to destroying discriminative regions in fine-grained scenarios. To address the"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.19956","kind":"arxiv","version":1},"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/2605.19956/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":"2605.19956","created_at":"2026-05-20T02:05:56.912965+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.19956v1","created_at":"2026-05-20T02:05:56.912965+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.19956","created_at":"2026-05-20T02:05:56.912965+00:00"},{"alias_kind":"pith_short_12","alias_value":"FDZWHDXGOITF","created_at":"2026-05-20T02:05:56.912965+00:00"},{"alias_kind":"pith_short_16","alias_value":"FDZWHDXGOITF4NBP","created_at":"2026-05-20T02:05:56.912965+00:00"},{"alias_kind":"pith_short_8","alias_value":"FDZWHDXG","created_at":"2026-05-20T02:05:56.912965+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/FDZWHDXGOITF4NBPZGAB7UNL2I","json":"https://pith.science/pith/FDZWHDXGOITF4NBPZGAB7UNL2I.json","graph_json":"https://pith.science/api/pith-number/FDZWHDXGOITF4NBPZGAB7UNL2I/graph.json","events_json":"https://pith.science/api/pith-number/FDZWHDXGOITF4NBPZGAB7UNL2I/events.json","paper":"https://pith.science/paper/FDZWHDXG"},"agent_actions":{"view_html":"https://pith.science/pith/FDZWHDXGOITF4NBPZGAB7UNL2I","download_json":"https://pith.science/pith/FDZWHDXGOITF4NBPZGAB7UNL2I.json","view_paper":"https://pith.science/paper/FDZWHDXG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.19956&json=true","fetch_graph":"https://pith.science/api/pith-number/FDZWHDXGOITF4NBPZGAB7UNL2I/graph.json","fetch_events":"https://pith.science/api/pith-number/FDZWHDXGOITF4NBPZGAB7UNL2I/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FDZWHDXGOITF4NBPZGAB7UNL2I/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FDZWHDXGOITF4NBPZGAB7UNL2I/action/storage_attestation","attest_author":"https://pith.science/pith/FDZWHDXGOITF4NBPZGAB7UNL2I/action/author_attestation","sign_citation":"https://pith.science/pith/FDZWHDXGOITF4NBPZGAB7UNL2I/action/citation_signature","submit_replication":"https://pith.science/pith/FDZWHDXGOITF4NBPZGAB7UNL2I/action/replication_record"}},"created_at":"2026-05-20T02:05:56.912965+00:00","updated_at":"2026-05-20T02:05:56.912965+00:00"}