{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:3A3AO5IFB67TOEYDKWCODFGQBL","short_pith_number":"pith:3A3AO5IF","schema_version":"1.0","canonical_sha256":"d8360775050fbf3713035584e194d00ad7a3cae6888e8388de24eeb368ae2c0c","source":{"kind":"arxiv","id":"2203.17274","version":2},"attestation_state":"computed","paper":{"title":"Exploring Visual Prompts for Adapting Large-Scale Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Ali Jahanian, Hyojin Bahng, Phillip Isola, Swami Sankaranarayanan","submitted_at":"2022-03-31T17:59:30Z","abstract_excerpt":"We investigate the efficacy of visual prompting to adapt large-scale models in vision. Following the recent approach from prompt tuning and adversarial reprogramming, we learn a single image perturbation such that a frozen model prompted with this perturbation performs a new task. Through comprehensive experiments, we demonstrate that visual prompting is particularly effective for CLIP and robust to distribution shift, achieving performance competitive with standard linear probes. We further analyze properties of the downstream dataset, prompt design, and output transformation in regard to ada"},"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":"2203.17274","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2022-03-31T17:59:30Z","cross_cats_sorted":[],"title_canon_sha256":"88b4600e6cbb954d80c1e1e26a5584e15fad52e12b3dda7a29854af0eef938bc","abstract_canon_sha256":"244319d4b2d0080270a387c33b9a2c1f2d3dc9ec5d836f69253e5e00eef88d99"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:28:43.832016Z","signature_b64":"aQ4UiW5tWJA6YSaxrx60s2RSEQfn+PDA6ALw9bH3N0gry5r0SdJW8cfmNh5LOx9j3zXjeU210GpJA3txdaeWDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d8360775050fbf3713035584e194d00ad7a3cae6888e8388de24eeb368ae2c0c","last_reissued_at":"2026-07-05T04:28:43.831564Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:28:43.831564Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Exploring Visual Prompts for Adapting Large-Scale Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Ali Jahanian, Hyojin Bahng, Phillip Isola, Swami Sankaranarayanan","submitted_at":"2022-03-31T17:59:30Z","abstract_excerpt":"We investigate the efficacy of visual prompting to adapt large-scale models in vision. Following the recent approach from prompt tuning and adversarial reprogramming, we learn a single image perturbation such that a frozen model prompted with this perturbation performs a new task. Through comprehensive experiments, we demonstrate that visual prompting is particularly effective for CLIP and robust to distribution shift, achieving performance competitive with standard linear probes. We further analyze properties of the downstream dataset, prompt design, and output transformation in regard to ada"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2203.17274","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/2203.17274/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":"2203.17274","created_at":"2026-07-05T04:28:43.831621+00:00"},{"alias_kind":"arxiv_version","alias_value":"2203.17274v2","created_at":"2026-07-05T04:28:43.831621+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2203.17274","created_at":"2026-07-05T04:28:43.831621+00:00"},{"alias_kind":"pith_short_12","alias_value":"3A3AO5IFB67T","created_at":"2026-07-05T04:28:43.831621+00:00"},{"alias_kind":"pith_short_16","alias_value":"3A3AO5IFB67TOEYD","created_at":"2026-07-05T04:28:43.831621+00:00"},{"alias_kind":"pith_short_8","alias_value":"3A3AO5IF","created_at":"2026-07-05T04:28:43.831621+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":14,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.19584","citing_title":"Language-Instructed Vision Embeddings for Controllable and Generalizable Perception","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11854","citing_title":"Fine-tuning Multi-modal LLMs with ART: Art-based Reinforcement Training","ref_index":13,"is_internal_anchor":false},{"citing_arxiv_id":"2605.31246","citing_title":"BadBone: Backdoor Attacks Against Backbone Models in Visual Prompt Learning","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2606.29600","citing_title":"One Scene, Two Depths: Probing Geometric Ambiguity in Monocular Foundation Models","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2606.00776","citing_title":"Latent Diffusion Pretraining for Crystal Property Prediction","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2402.10380","citing_title":"Subgraph-level Universal Prompt Tuning","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2407.17491","citing_title":"Robust Adaptation of Foundation Models with Black-Box Visual Prompting","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17577","citing_title":"TAME: Test-Time Adversarial Prompt Tuning via Mixture-of-Experts for Vision-Language Models","ref_index":66,"is_internal_anchor":false},{"citing_arxiv_id":"2402.07927","citing_title":"A Systematic Survey of Prompt Engineering in Large Language Models: Techniques and Applications","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08273","citing_title":"Efficient Prompt Learning for Traffic Forecasting","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10130","citing_title":"Thermal-Det: Language-Guided Cross-Modal Distillation for Open-Vocabulary Thermal Object Detection","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05910","citing_title":"Plug-and-play Class-aware Knowledge Injection for Prompt Learning with Visual-Language Model","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00906","citing_title":"Generalized Category Discovery under Domain Shifts: From Vision to Vision-Language Models","ref_index":69,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06440","citing_title":"Visual prompting reimagined: The power of the Activation Prompts","ref_index":50,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3A3AO5IFB67TOEYDKWCODFGQBL","json":"https://pith.science/pith/3A3AO5IFB67TOEYDKWCODFGQBL.json","graph_json":"https://pith.science/api/pith-number/3A3AO5IFB67TOEYDKWCODFGQBL/graph.json","events_json":"https://pith.science/api/pith-number/3A3AO5IFB67TOEYDKWCODFGQBL/events.json","paper":"https://pith.science/paper/3A3AO5IF"},"agent_actions":{"view_html":"https://pith.science/pith/3A3AO5IFB67TOEYDKWCODFGQBL","download_json":"https://pith.science/pith/3A3AO5IFB67TOEYDKWCODFGQBL.json","view_paper":"https://pith.science/paper/3A3AO5IF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2203.17274&json=true","fetch_graph":"https://pith.science/api/pith-number/3A3AO5IFB67TOEYDKWCODFGQBL/graph.json","fetch_events":"https://pith.science/api/pith-number/3A3AO5IFB67TOEYDKWCODFGQBL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3A3AO5IFB67TOEYDKWCODFGQBL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3A3AO5IFB67TOEYDKWCODFGQBL/action/storage_attestation","attest_author":"https://pith.science/pith/3A3AO5IFB67TOEYDKWCODFGQBL/action/author_attestation","sign_citation":"https://pith.science/pith/3A3AO5IFB67TOEYDKWCODFGQBL/action/citation_signature","submit_replication":"https://pith.science/pith/3A3AO5IFB67TOEYDKWCODFGQBL/action/replication_record"}},"created_at":"2026-07-05T04:28:43.831621+00:00","updated_at":"2026-07-05T04:28:43.831621+00:00"}