{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:LSJXXDE7ECFK5XGWHVGBKKX77Y","short_pith_number":"pith:LSJXXDE7","schema_version":"1.0","canonical_sha256":"5c937b8c9f208aaedcd63d4c152afffe219760c4c30f9826392e6fde681ae52b","source":{"kind":"arxiv","id":"2203.12119","version":2},"attestation_state":"computed","paper":{"title":"Visual Prompt Tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bharath Hariharan, Bor-Chun Chen, Claire Cardie, Luming Tang, Menglin Jia, Serge Belongie, Ser-Nam Lim","submitted_at":"2022-03-23T01:17:16Z","abstract_excerpt":"The current modus operandi in adapting pre-trained models involves updating all the backbone parameters, ie, full fine-tuning. This paper introduces Visual Prompt Tuning (VPT) as an efficient and effective alternative to full fine-tuning for large-scale Transformer models in vision. Taking inspiration from recent advances in efficiently tuning large language models, VPT introduces only a small amount (less than 1% of model parameters) of trainable parameters in the input space while keeping the model backbone frozen. Via extensive experiments on a wide variety of downstream recognition tasks, "},"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":"2203.12119","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2022-03-23T01:17:16Z","cross_cats_sorted":[],"title_canon_sha256":"ae358d10a65eabb0ef4312d2baf1f800f3a72a9327ed0b6c281bbccf62edb81b","abstract_canon_sha256":"0ae41cf4894456a89e06bf90b408033ae62d99d60ad2fad2afaa4b6124907683"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:42:09.289343Z","signature_b64":"Y8Wfwyw4kFRN/GJ8Ynxcx+6RG3mKH7Pswa0MCT/S556OQ/vf6PDzDY5I8vYHdEPtjaae2weDsPakbTL4Ahi2AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5c937b8c9f208aaedcd63d4c152afffe219760c4c30f9826392e6fde681ae52b","last_reissued_at":"2026-07-05T04:42:09.288848Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:42:09.288848Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Visual Prompt Tuning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bharath Hariharan, Bor-Chun Chen, Claire Cardie, Luming Tang, Menglin Jia, Serge Belongie, Ser-Nam Lim","submitted_at":"2022-03-23T01:17:16Z","abstract_excerpt":"The current modus operandi in adapting pre-trained models involves updating all the backbone parameters, ie, full fine-tuning. This paper introduces Visual Prompt Tuning (VPT) as an efficient and effective alternative to full fine-tuning for large-scale Transformer models in vision. Taking inspiration from recent advances in efficiently tuning large language models, VPT introduces only a small amount (less than 1% of model parameters) of trainable parameters in the input space while keeping the model backbone frozen. Via extensive experiments on a wide variety of downstream recognition tasks, "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2203.12119","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/2203.12119/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":"2203.12119","created_at":"2026-07-05T04:42:09.288920+00:00"},{"alias_kind":"arxiv_version","alias_value":"2203.12119v2","created_at":"2026-07-05T04:42:09.288920+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2203.12119","created_at":"2026-07-05T04:42:09.288920+00:00"},{"alias_kind":"pith_short_12","alias_value":"LSJXXDE7ECFK","created_at":"2026-07-05T04:42:09.288920+00:00"},{"alias_kind":"pith_short_16","alias_value":"LSJXXDE7ECFK5XGW","created_at":"2026-07-05T04:42:09.288920+00:00"},{"alias_kind":"pith_short_8","alias_value":"LSJXXDE7","created_at":"2026-07-05T04:42:09.288920+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.11854","citing_title":"Fine-tuning Multi-modal LLMs with ART: Art-based Reinforcement Training","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2407.17491","citing_title":"Robust Adaptation of Foundation Models with Black-Box Visual Prompting","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06440","citing_title":"Visual prompting reimagined: The power of the Activation Prompts","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2410.02713","citing_title":"LLaVA-Video: Video Instruction Tuning With Synthetic Data","ref_index":74,"is_internal_anchor":false},{"citing_arxiv_id":"2605.01048","citing_title":"Compared to What? Baselines and Metrics for Counterfactual Prompting","ref_index":96,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LSJXXDE7ECFK5XGWHVGBKKX77Y","json":"https://pith.science/pith/LSJXXDE7ECFK5XGWHVGBKKX77Y.json","graph_json":"https://pith.science/api/pith-number/LSJXXDE7ECFK5XGWHVGBKKX77Y/graph.json","events_json":"https://pith.science/api/pith-number/LSJXXDE7ECFK5XGWHVGBKKX77Y/events.json","paper":"https://pith.science/paper/LSJXXDE7"},"agent_actions":{"view_html":"https://pith.science/pith/LSJXXDE7ECFK5XGWHVGBKKX77Y","download_json":"https://pith.science/pith/LSJXXDE7ECFK5XGWHVGBKKX77Y.json","view_paper":"https://pith.science/paper/LSJXXDE7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2203.12119&json=true","fetch_graph":"https://pith.science/api/pith-number/LSJXXDE7ECFK5XGWHVGBKKX77Y/graph.json","fetch_events":"https://pith.science/api/pith-number/LSJXXDE7ECFK5XGWHVGBKKX77Y/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LSJXXDE7ECFK5XGWHVGBKKX77Y/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LSJXXDE7ECFK5XGWHVGBKKX77Y/action/storage_attestation","attest_author":"https://pith.science/pith/LSJXXDE7ECFK5XGWHVGBKKX77Y/action/author_attestation","sign_citation":"https://pith.science/pith/LSJXXDE7ECFK5XGWHVGBKKX77Y/action/citation_signature","submit_replication":"https://pith.science/pith/LSJXXDE7ECFK5XGWHVGBKKX77Y/action/replication_record"}},"created_at":"2026-07-05T04:42:09.288920+00:00","updated_at":"2026-07-05T04:42:09.288920+00:00"}