{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:VLIURWWPU3LV4KKGCYLU3NGDC6","short_pith_number":"pith:VLIURWWP","schema_version":"1.0","canonical_sha256":"aad148dacfa6d75e294616174db4c317a0463577c6578f5b0c18e7e59d5d0eef","source":{"kind":"arxiv","id":"2410.13121","version":1},"attestation_state":"computed","paper":{"title":"Trust but Verify: Programmatic VLM Evaluation in the Wild","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"An Yan, Caiming Xiong, Ran Xu, Senthil Purushwalkam, Viraj Prabhu","submitted_at":"2024-10-17T01:19:18Z","abstract_excerpt":"Vision-Language Models (VLMs) often generate plausible but incorrect responses to visual queries. However, reliably quantifying the effect of such hallucinations in free-form responses to open-ended queries is challenging as it requires visually verifying each claim within the response. We propose Programmatic VLM Evaluation (PROVE), a new benchmarking paradigm for evaluating VLM responses to open-ended queries. To construct PROVE, we provide a large language model (LLM) with a high-fidelity scene-graph representation constructed from a hyper-detailed image caption, and prompt it to generate d"},"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":"2410.13121","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-10-17T01:19:18Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"84bb91261ce46c8d7ad175fe17c8d9430357fd7c0aa1eace9421f6996ee368cf","abstract_canon_sha256":"317b2681e5c99a11bceb6c96ed2797904ced183f6cab0af5ddd672b03f6eaaf9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:21:55.858479Z","signature_b64":"RXfGWNOiUhwzLLeD7M84dWkG9hFd0R9LnUdM/JKZ660N5Epw4sqxo9WJBLPOFxjaky4syzKrNdDuzZszfaRrAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"aad148dacfa6d75e294616174db4c317a0463577c6578f5b0c18e7e59d5d0eef","last_reissued_at":"2026-07-05T09:21:55.857909Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:21:55.857909Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Trust but Verify: Programmatic VLM Evaluation in the Wild","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"An Yan, Caiming Xiong, Ran Xu, Senthil Purushwalkam, Viraj Prabhu","submitted_at":"2024-10-17T01:19:18Z","abstract_excerpt":"Vision-Language Models (VLMs) often generate plausible but incorrect responses to visual queries. However, reliably quantifying the effect of such hallucinations in free-form responses to open-ended queries is challenging as it requires visually verifying each claim within the response. We propose Programmatic VLM Evaluation (PROVE), a new benchmarking paradigm for evaluating VLM responses to open-ended queries. To construct PROVE, we provide a large language model (LLM) with a high-fidelity scene-graph representation constructed from a hyper-detailed image caption, and prompt it to generate d"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.13121","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/2410.13121/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":"2410.13121","created_at":"2026-07-05T09:21:55.857987+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.13121v1","created_at":"2026-07-05T09:21:55.857987+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.13121","created_at":"2026-07-05T09:21:55.857987+00:00"},{"alias_kind":"pith_short_12","alias_value":"VLIURWWPU3LV","created_at":"2026-07-05T09:21:55.857987+00:00"},{"alias_kind":"pith_short_16","alias_value":"VLIURWWPU3LV4KKG","created_at":"2026-07-05T09:21:55.857987+00:00"},{"alias_kind":"pith_short_8","alias_value":"VLIURWWP","created_at":"2026-07-05T09:21:55.857987+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.24073","citing_title":"mRAG: Elucidating the Design Space of Multi-modal Retrieval-Augmented Generation","ref_index":35,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VLIURWWPU3LV4KKGCYLU3NGDC6","json":"https://pith.science/pith/VLIURWWPU3LV4KKGCYLU3NGDC6.json","graph_json":"https://pith.science/api/pith-number/VLIURWWPU3LV4KKGCYLU3NGDC6/graph.json","events_json":"https://pith.science/api/pith-number/VLIURWWPU3LV4KKGCYLU3NGDC6/events.json","paper":"https://pith.science/paper/VLIURWWP"},"agent_actions":{"view_html":"https://pith.science/pith/VLIURWWPU3LV4KKGCYLU3NGDC6","download_json":"https://pith.science/pith/VLIURWWPU3LV4KKGCYLU3NGDC6.json","view_paper":"https://pith.science/paper/VLIURWWP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.13121&json=true","fetch_graph":"https://pith.science/api/pith-number/VLIURWWPU3LV4KKGCYLU3NGDC6/graph.json","fetch_events":"https://pith.science/api/pith-number/VLIURWWPU3LV4KKGCYLU3NGDC6/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VLIURWWPU3LV4KKGCYLU3NGDC6/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VLIURWWPU3LV4KKGCYLU3NGDC6/action/storage_attestation","attest_author":"https://pith.science/pith/VLIURWWPU3LV4KKGCYLU3NGDC6/action/author_attestation","sign_citation":"https://pith.science/pith/VLIURWWPU3LV4KKGCYLU3NGDC6/action/citation_signature","submit_replication":"https://pith.science/pith/VLIURWWPU3LV4KKGCYLU3NGDC6/action/replication_record"}},"created_at":"2026-07-05T09:21:55.857987+00:00","updated_at":"2026-07-05T09:21:55.857987+00:00"}