{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:VEJ4LBZ4RMCQ3RTCMPMQEDBI6M","short_pith_number":"pith:VEJ4LBZ4","schema_version":"1.0","canonical_sha256":"a913c5873c8b050dc66263d9020c28f32aedb6ae239ed6f4ce0cf34023adc92b","source":{"kind":"arxiv","id":"2506.05146","version":2},"attestation_state":"computed","paper":{"title":"CIVET: Systematic Evaluation of Understanding in VLMs","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Gabriel Roccabruna, Giuseppe Riccardi, Massimo Rizzoli, Olha Khomyn, Seyed Mahed Mousavi, Simone Alghisi","submitted_at":"2025-06-05T15:27:16Z","abstract_excerpt":"While Vision-Language Models (VLMs) have achieved competitive performance in various tasks, their comprehension of the underlying structure and semantics of a scene remains understudied. To investigate the understanding of VLMs, we study their capability regarding object properties and relations in a controlled and interpretable manner. To this scope, we introduce CIVET, a novel and extensible framework for systematiC evaluatIon Via controllEd sTimuli. CIVET addresses the lack of standardized systematic evaluation for assessing VLMs' understanding, enabling researchers to test hypotheses with "},"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":"2506.05146","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2025-06-05T15:27:16Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"d9b049cebc82350ba330cff99dd707b580d3f75c0c739fb75351f9c6aaecd011","abstract_canon_sha256":"9d0ba279b38a52df0be0d1d81c72f610d26e480e46684c9ecb166a9649a1167d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:24:22.112988Z","signature_b64":"Ot7Soyt69Eq0TrZwMj/ZNo0BD5siJI/O5G3wg6SugcktWqJrqFHspxCGLoI47+7EyKD+T/FUxpip6g87sa53Aw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a913c5873c8b050dc66263d9020c28f32aedb6ae239ed6f4ce0cf34023adc92b","last_reissued_at":"2026-07-05T11:24:22.112481Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:24:22.112481Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CIVET: Systematic Evaluation of Understanding in VLMs","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Gabriel Roccabruna, Giuseppe Riccardi, Massimo Rizzoli, Olha Khomyn, Seyed Mahed Mousavi, Simone Alghisi","submitted_at":"2025-06-05T15:27:16Z","abstract_excerpt":"While Vision-Language Models (VLMs) have achieved competitive performance in various tasks, their comprehension of the underlying structure and semantics of a scene remains understudied. To investigate the understanding of VLMs, we study their capability regarding object properties and relations in a controlled and interpretable manner. To this scope, we introduce CIVET, a novel and extensible framework for systematiC evaluatIon Via controllEd sTimuli. CIVET addresses the lack of standardized systematic evaluation for assessing VLMs' understanding, enabling researchers to test hypotheses with "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.05146","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/2506.05146/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":"2506.05146","created_at":"2026-07-05T11:24:22.112542+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.05146v2","created_at":"2026-07-05T11:24:22.112542+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.05146","created_at":"2026-07-05T11:24:22.112542+00:00"},{"alias_kind":"pith_short_12","alias_value":"VEJ4LBZ4RMCQ","created_at":"2026-07-05T11:24:22.112542+00:00"},{"alias_kind":"pith_short_16","alias_value":"VEJ4LBZ4RMCQ3RTC","created_at":"2026-07-05T11:24:22.112542+00:00"},{"alias_kind":"pith_short_8","alias_value":"VEJ4LBZ4","created_at":"2026-07-05T11:24:22.112542+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/VEJ4LBZ4RMCQ3RTCMPMQEDBI6M","json":"https://pith.science/pith/VEJ4LBZ4RMCQ3RTCMPMQEDBI6M.json","graph_json":"https://pith.science/api/pith-number/VEJ4LBZ4RMCQ3RTCMPMQEDBI6M/graph.json","events_json":"https://pith.science/api/pith-number/VEJ4LBZ4RMCQ3RTCMPMQEDBI6M/events.json","paper":"https://pith.science/paper/VEJ4LBZ4"},"agent_actions":{"view_html":"https://pith.science/pith/VEJ4LBZ4RMCQ3RTCMPMQEDBI6M","download_json":"https://pith.science/pith/VEJ4LBZ4RMCQ3RTCMPMQEDBI6M.json","view_paper":"https://pith.science/paper/VEJ4LBZ4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.05146&json=true","fetch_graph":"https://pith.science/api/pith-number/VEJ4LBZ4RMCQ3RTCMPMQEDBI6M/graph.json","fetch_events":"https://pith.science/api/pith-number/VEJ4LBZ4RMCQ3RTCMPMQEDBI6M/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VEJ4LBZ4RMCQ3RTCMPMQEDBI6M/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VEJ4LBZ4RMCQ3RTCMPMQEDBI6M/action/storage_attestation","attest_author":"https://pith.science/pith/VEJ4LBZ4RMCQ3RTCMPMQEDBI6M/action/author_attestation","sign_citation":"https://pith.science/pith/VEJ4LBZ4RMCQ3RTCMPMQEDBI6M/action/citation_signature","submit_replication":"https://pith.science/pith/VEJ4LBZ4RMCQ3RTCMPMQEDBI6M/action/replication_record"}},"created_at":"2026-07-05T11:24:22.112542+00:00","updated_at":"2026-07-05T11:24:22.112542+00:00"}