{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:NCVUURDFKNQYQGWFSFJK2OMHQ4","short_pith_number":"pith:NCVUURDF","schema_version":"1.0","canonical_sha256":"68ab4a44655361881ac59152ad39878711a6dc4d5506fcedc6e6711da39b0518","source":{"kind":"arxiv","id":"2507.00045","version":1},"attestation_state":"computed","paper":{"title":"CaughtCheating: Is Your MLLM a Good Cheating Detective? Exploring the Boundary of Visual Perception and Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CV","authors_text":"Chenguang Wang, Ming Li, Ruiyi Zhang, Tianyi Zhou, Xiyang Wu, Xiyao Wang, Yijun Liang, Yuhang Zhou, Yuqing Zhang","submitted_at":"2025-06-23T22:05:21Z","abstract_excerpt":"Recent agentic Multi-Modal Large Language Models (MLLMs) such as GPT-o3 have achieved near-ceiling scores on various existing benchmarks, motivating a demand for more challenging test tasks. These MLLMs have been reported to excel in a few expert-level tasks for humans, e.g., GeoGuesser, reflecting their potential as a detective who can notice minuscule cues in an image and weave them into coherent, situational explanations, leading to a reliable answer. But can they match the performance of excellent human detectives? To answer this question, we investigate some hard scenarios where GPT-o3 ca"},"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":"2507.00045","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-06-23T22:05:21Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"a36cc468cebae109a4973e776762580c668b6d66d8fea29b90e66bf8b881463a","abstract_canon_sha256":"19fe163bad4fd265f850c5aa50e57578ca7ae373cdb10d4838e6f3bcf1363055"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:29:35.644261Z","signature_b64":"ICPGu/Gcl6gN6eKTUbcz2RbPWHiZjXuAdVgQAsiAj/ICy1TL2HMg9dJeJGGj3LMJZMubi/A1MIvsyeOoJVx8Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"68ab4a44655361881ac59152ad39878711a6dc4d5506fcedc6e6711da39b0518","last_reissued_at":"2026-07-05T11:29:35.643601Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:29:35.643601Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"CaughtCheating: Is Your MLLM a Good Cheating Detective? Exploring the Boundary of Visual Perception and Reasoning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.CV","authors_text":"Chenguang Wang, Ming Li, Ruiyi Zhang, Tianyi Zhou, Xiyang Wu, Xiyao Wang, Yijun Liang, Yuhang Zhou, Yuqing Zhang","submitted_at":"2025-06-23T22:05:21Z","abstract_excerpt":"Recent agentic Multi-Modal Large Language Models (MLLMs) such as GPT-o3 have achieved near-ceiling scores on various existing benchmarks, motivating a demand for more challenging test tasks. These MLLMs have been reported to excel in a few expert-level tasks for humans, e.g., GeoGuesser, reflecting their potential as a detective who can notice minuscule cues in an image and weave them into coherent, situational explanations, leading to a reliable answer. But can they match the performance of excellent human detectives? To answer this question, we investigate some hard scenarios where GPT-o3 ca"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2507.00045","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/2507.00045/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":"2507.00045","created_at":"2026-07-05T11:29:35.643696+00:00"},{"alias_kind":"arxiv_version","alias_value":"2507.00045v1","created_at":"2026-07-05T11:29:35.643696+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2507.00045","created_at":"2026-07-05T11:29:35.643696+00:00"},{"alias_kind":"pith_short_12","alias_value":"NCVUURDFKNQY","created_at":"2026-07-05T11:29:35.643696+00:00"},{"alias_kind":"pith_short_16","alias_value":"NCVUURDFKNQYQGWF","created_at":"2026-07-05T11:29:35.643696+00:00"},{"alias_kind":"pith_short_8","alias_value":"NCVUURDF","created_at":"2026-07-05T11:29:35.643696+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2508.07493","citing_title":"VisR-Bench: An Empirical Study on Visual Retrieval-Augmented Generation for Multilingual Long Document Understanding","ref_index":21,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NCVUURDFKNQYQGWFSFJK2OMHQ4","json":"https://pith.science/pith/NCVUURDFKNQYQGWFSFJK2OMHQ4.json","graph_json":"https://pith.science/api/pith-number/NCVUURDFKNQYQGWFSFJK2OMHQ4/graph.json","events_json":"https://pith.science/api/pith-number/NCVUURDFKNQYQGWFSFJK2OMHQ4/events.json","paper":"https://pith.science/paper/NCVUURDF"},"agent_actions":{"view_html":"https://pith.science/pith/NCVUURDFKNQYQGWFSFJK2OMHQ4","download_json":"https://pith.science/pith/NCVUURDFKNQYQGWFSFJK2OMHQ4.json","view_paper":"https://pith.science/paper/NCVUURDF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2507.00045&json=true","fetch_graph":"https://pith.science/api/pith-number/NCVUURDFKNQYQGWFSFJK2OMHQ4/graph.json","fetch_events":"https://pith.science/api/pith-number/NCVUURDFKNQYQGWFSFJK2OMHQ4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NCVUURDFKNQYQGWFSFJK2OMHQ4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NCVUURDFKNQYQGWFSFJK2OMHQ4/action/storage_attestation","attest_author":"https://pith.science/pith/NCVUURDFKNQYQGWFSFJK2OMHQ4/action/author_attestation","sign_citation":"https://pith.science/pith/NCVUURDFKNQYQGWFSFJK2OMHQ4/action/citation_signature","submit_replication":"https://pith.science/pith/NCVUURDFKNQYQGWFSFJK2OMHQ4/action/replication_record"}},"created_at":"2026-07-05T11:29:35.643696+00:00","updated_at":"2026-07-05T11:29:35.643696+00:00"}