{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:3EIDTBYIGG5U5KPSOYGG33JI62","short_pith_number":"pith:3EIDTBYI","schema_version":"1.0","canonical_sha256":"d91039870831bb4ea9f2760c6ded28f6a323a53c9010dc5abcbb82b30924fe36","source":{"kind":"arxiv","id":"2507.15480","version":2},"attestation_state":"computed","paper":{"title":"One Last Attention for Your Vision-Language Model","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Ghazi Shazan Ahmad, Liang Chen, Lingqiao Liu, Tianjun Yao, Zhiqiang Shen","submitted_at":"2025-07-21T10:35:32Z","abstract_excerpt":"Pretrained vision-language models (VLMs), such as CLIP, achieve remarkable zero-shot performance, yet their downstream potential hinges on effective fine-tuning. Most adaptation methods typically focus on refining representation from separate modalities (text or vision) but neglect the critical role of their fused representations in the decision-making process, \\emph{\\ie} rational matrix that drives the final prediction. To bridge the gap, we propose a simple yet effective \\textbf{R}ational \\textbf{Ada}ptaion ({RAda}) to explicitly exploit the final fused representation during fine-tuning. RAd"},"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":"2507.15480","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-07-21T10:35:32Z","cross_cats_sorted":[],"title_canon_sha256":"6b3b9946e7cc1c5c5882e79bc6f9c788571bb748806e1ef9ea560880a00586ba","abstract_canon_sha256":"a88f77ced9ce11421392834a9d498fd125c5295acbd0b9830bf02769c37bfa90"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:44:09.474191Z","signature_b64":"nQBjgXzKaCLGkMA+r+2M2FmOgmYKlhmecaMGHgzTWwLmNUg3v9ZjZmV191m7Id7c41R5kO+7qDJvC4jM4v5VBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d91039870831bb4ea9f2760c6ded28f6a323a53c9010dc5abcbb82b30924fe36","last_reissued_at":"2026-07-05T11:44:09.473672Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:44:09.473672Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"One Last Attention for Your Vision-Language Model","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Ghazi Shazan Ahmad, Liang Chen, Lingqiao Liu, Tianjun Yao, Zhiqiang Shen","submitted_at":"2025-07-21T10:35:32Z","abstract_excerpt":"Pretrained vision-language models (VLMs), such as CLIP, achieve remarkable zero-shot performance, yet their downstream potential hinges on effective fine-tuning. Most adaptation methods typically focus on refining representation from separate modalities (text or vision) but neglect the critical role of their fused representations in the decision-making process, \\emph{\\ie} rational matrix that drives the final prediction. To bridge the gap, we propose a simple yet effective \\textbf{R}ational \\textbf{Ada}ptaion ({RAda}) to explicitly exploit the final fused representation during fine-tuning. RAd"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.15480","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/2507.15480/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":"2507.15480","created_at":"2026-07-05T11:44:09.473737+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.15480v2","created_at":"2026-07-05T11:44:09.473737+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.15480","created_at":"2026-07-05T11:44:09.473737+00:00"},{"alias_kind":"pith_short_12","alias_value":"3EIDTBYIGG5U","created_at":"2026-07-05T11:44:09.473737+00:00"},{"alias_kind":"pith_short_16","alias_value":"3EIDTBYIGG5U5KPS","created_at":"2026-07-05T11:44:09.473737+00:00"},{"alias_kind":"pith_short_8","alias_value":"3EIDTBYI","created_at":"2026-07-05T11:44:09.473737+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/3EIDTBYIGG5U5KPSOYGG33JI62","json":"https://pith.science/pith/3EIDTBYIGG5U5KPSOYGG33JI62.json","graph_json":"https://pith.science/api/pith-number/3EIDTBYIGG5U5KPSOYGG33JI62/graph.json","events_json":"https://pith.science/api/pith-number/3EIDTBYIGG5U5KPSOYGG33JI62/events.json","paper":"https://pith.science/paper/3EIDTBYI"},"agent_actions":{"view_html":"https://pith.science/pith/3EIDTBYIGG5U5KPSOYGG33JI62","download_json":"https://pith.science/pith/3EIDTBYIGG5U5KPSOYGG33JI62.json","view_paper":"https://pith.science/paper/3EIDTBYI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.15480&json=true","fetch_graph":"https://pith.science/api/pith-number/3EIDTBYIGG5U5KPSOYGG33JI62/graph.json","fetch_events":"https://pith.science/api/pith-number/3EIDTBYIGG5U5KPSOYGG33JI62/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3EIDTBYIGG5U5KPSOYGG33JI62/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3EIDTBYIGG5U5KPSOYGG33JI62/action/storage_attestation","attest_author":"https://pith.science/pith/3EIDTBYIGG5U5KPSOYGG33JI62/action/author_attestation","sign_citation":"https://pith.science/pith/3EIDTBYIGG5U5KPSOYGG33JI62/action/citation_signature","submit_replication":"https://pith.science/pith/3EIDTBYIGG5U5KPSOYGG33JI62/action/replication_record"}},"created_at":"2026-07-05T11:44:09.473737+00:00","updated_at":"2026-07-05T11:44:09.473737+00:00"}