{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:AIR5VAUOI63XGEEZW6K74BIFEN","short_pith_number":"pith:AIR5VAUO","schema_version":"1.0","canonical_sha256":"0223da828e47b7731099b795fe0505234c9591cefbad6ff633196d4d9bbd7654","source":{"kind":"arxiv","id":"2208.08831","version":2},"attestation_state":"computed","paper":{"title":"Discovering Bugs in Vision Models using Off-the-shelf Image Generation and Captioning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.CV","authors_text":"Isabela Albuquerque, Olivia Wiles, Sven Gowal","submitted_at":"2022-08-18T13:49:10Z","abstract_excerpt":"Automatically discovering failures in vision models under real-world settings remains an open challenge. This work demonstrates how off-the-shelf, large-scale, image-to-text and text-to-image models, trained on vast amounts of data, can be leveraged to automatically find such failures. In essence, a conditional text-to-image generative model is used to generate large amounts of synthetic, yet realistic, inputs given a ground-truth label. Misclassified inputs are clustered and a captioning model is used to describe each cluster. Each cluster's description is used in turn to generate more inputs"},"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":"2208.08831","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2022-08-18T13:49:10Z","cross_cats_sorted":["cs.LG","stat.ML"],"title_canon_sha256":"b81b7c74e16df46dad8e6fb78f2d643336e24fddaa855e7a44a464fdbf423e65","abstract_canon_sha256":"19212c361b8c67d6ab8e8dac9d0a19281580b0423540ff1f8da67afa0762cd53"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:09:05.326114Z","signature_b64":"ib4IYWXVN9zVNCcVeawK3CZGbF2ZJ1UtewdSwKcTi0PVwqcboNsuHO7DAiJqESiGLRiyiHDsUW5QjCUgoAZSCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0223da828e47b7731099b795fe0505234c9591cefbad6ff633196d4d9bbd7654","last_reissued_at":"2026-07-05T06:09:05.325643Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:09:05.325643Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Discovering Bugs in Vision Models using Off-the-shelf Image Generation and Captioning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","stat.ML"],"primary_cat":"cs.CV","authors_text":"Isabela Albuquerque, Olivia Wiles, Sven Gowal","submitted_at":"2022-08-18T13:49:10Z","abstract_excerpt":"Automatically discovering failures in vision models under real-world settings remains an open challenge. This work demonstrates how off-the-shelf, large-scale, image-to-text and text-to-image models, trained on vast amounts of data, can be leveraged to automatically find such failures. In essence, a conditional text-to-image generative model is used to generate large amounts of synthetic, yet realistic, inputs given a ground-truth label. Misclassified inputs are clustered and a captioning model is used to describe each cluster. Each cluster's description is used in turn to generate more inputs"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2208.08831","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/2208.08831/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":"2208.08831","created_at":"2026-07-05T06:09:05.325700+00:00"},{"alias_kind":"arxiv_version","alias_value":"2208.08831v2","created_at":"2026-07-05T06:09:05.325700+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2208.08831","created_at":"2026-07-05T06:09:05.325700+00:00"},{"alias_kind":"pith_short_12","alias_value":"AIR5VAUOI63X","created_at":"2026-07-05T06:09:05.325700+00:00"},{"alias_kind":"pith_short_16","alias_value":"AIR5VAUOI63XGEEZ","created_at":"2026-07-05T06:09:05.325700+00:00"},{"alias_kind":"pith_short_8","alias_value":"AIR5VAUO","created_at":"2026-07-05T06:09:05.325700+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/AIR5VAUOI63XGEEZW6K74BIFEN","json":"https://pith.science/pith/AIR5VAUOI63XGEEZW6K74BIFEN.json","graph_json":"https://pith.science/api/pith-number/AIR5VAUOI63XGEEZW6K74BIFEN/graph.json","events_json":"https://pith.science/api/pith-number/AIR5VAUOI63XGEEZW6K74BIFEN/events.json","paper":"https://pith.science/paper/AIR5VAUO"},"agent_actions":{"view_html":"https://pith.science/pith/AIR5VAUOI63XGEEZW6K74BIFEN","download_json":"https://pith.science/pith/AIR5VAUOI63XGEEZW6K74BIFEN.json","view_paper":"https://pith.science/paper/AIR5VAUO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2208.08831&json=true","fetch_graph":"https://pith.science/api/pith-number/AIR5VAUOI63XGEEZW6K74BIFEN/graph.json","fetch_events":"https://pith.science/api/pith-number/AIR5VAUOI63XGEEZW6K74BIFEN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AIR5VAUOI63XGEEZW6K74BIFEN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AIR5VAUOI63XGEEZW6K74BIFEN/action/storage_attestation","attest_author":"https://pith.science/pith/AIR5VAUOI63XGEEZW6K74BIFEN/action/author_attestation","sign_citation":"https://pith.science/pith/AIR5VAUOI63XGEEZW6K74BIFEN/action/citation_signature","submit_replication":"https://pith.science/pith/AIR5VAUOI63XGEEZW6K74BIFEN/action/replication_record"}},"created_at":"2026-07-05T06:09:05.325700+00:00","updated_at":"2026-07-05T06:09:05.325700+00:00"}