{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:BFIWE5POTHK4TZEPD27QE5DGFN","short_pith_number":"pith:BFIWE5PO","schema_version":"1.0","canonical_sha256":"09516275ee99d5c9e48f1ebf0274662b42e47ec01b97d8d4d47b87d9fe5c8006","source":{"kind":"arxiv","id":"2505.22112","version":1},"attestation_state":"computed","paper":{"title":"Visual Large Language Models Exhibit Human-Level Cognitive Flexibility in the Wisconsin Card Sorting Test","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["q-bio.NC"],"primary_cat":"cs.AI","authors_text":"Frederic Alexandre, Guangfu Hao, Shan Yu","submitted_at":"2025-05-28T08:40:55Z","abstract_excerpt":"Cognitive flexibility has been extensively studied in human cognition but remains relatively unexplored in the context of Visual Large Language Models (VLLMs). This study assesses the cognitive flexibility of state-of-the-art VLLMs (GPT-4o, Gemini-1.5 Pro, and Claude-3.5 Sonnet) using the Wisconsin Card Sorting Test (WCST), a classic measure of set-shifting ability. Our results reveal that VLLMs achieve or surpass human-level set-shifting capabilities under chain-of-thought prompting with text-based inputs. However, their abilities are highly influenced by both input modality and prompting str"},"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":"2505.22112","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2025-05-28T08:40:55Z","cross_cats_sorted":["q-bio.NC"],"title_canon_sha256":"84c445337eea203fdd0339299790100c414a539aa823c464c9cfd3f4947ca3b6","abstract_canon_sha256":"533afd6e2d102c0bcdaca67fbd985d706649bfc47b6a9ae5b7c2084c67aa0b1a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:11:10.087461Z","signature_b64":"uLsVMlpogNzyxjc7mjoMIrO+BeuPs036Q9aD/4/E4ftDSY9MnEQyDLoDJEnX9jtUHGJOU8olA3+rnsGtpR2bAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"09516275ee99d5c9e48f1ebf0274662b42e47ec01b97d8d4d47b87d9fe5c8006","last_reissued_at":"2026-07-05T11:11:10.086956Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:11:10.086956Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Visual Large Language Models Exhibit Human-Level Cognitive Flexibility in the Wisconsin Card Sorting Test","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["q-bio.NC"],"primary_cat":"cs.AI","authors_text":"Frederic Alexandre, Guangfu Hao, Shan Yu","submitted_at":"2025-05-28T08:40:55Z","abstract_excerpt":"Cognitive flexibility has been extensively studied in human cognition but remains relatively unexplored in the context of Visual Large Language Models (VLLMs). This study assesses the cognitive flexibility of state-of-the-art VLLMs (GPT-4o, Gemini-1.5 Pro, and Claude-3.5 Sonnet) using the Wisconsin Card Sorting Test (WCST), a classic measure of set-shifting ability. Our results reveal that VLLMs achieve or surpass human-level set-shifting capabilities under chain-of-thought prompting with text-based inputs. However, their abilities are highly influenced by both input modality and prompting str"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.22112","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/2505.22112/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":"2505.22112","created_at":"2026-07-05T11:11:10.087015+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.22112v1","created_at":"2026-07-05T11:11:10.087015+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.22112","created_at":"2026-07-05T11:11:10.087015+00:00"},{"alias_kind":"pith_short_12","alias_value":"BFIWE5POTHK4","created_at":"2026-07-05T11:11:10.087015+00:00"},{"alias_kind":"pith_short_16","alias_value":"BFIWE5POTHK4TZEP","created_at":"2026-07-05T11:11:10.087015+00:00"},{"alias_kind":"pith_short_8","alias_value":"BFIWE5PO","created_at":"2026-07-05T11:11:10.087015+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/BFIWE5POTHK4TZEPD27QE5DGFN","json":"https://pith.science/pith/BFIWE5POTHK4TZEPD27QE5DGFN.json","graph_json":"https://pith.science/api/pith-number/BFIWE5POTHK4TZEPD27QE5DGFN/graph.json","events_json":"https://pith.science/api/pith-number/BFIWE5POTHK4TZEPD27QE5DGFN/events.json","paper":"https://pith.science/paper/BFIWE5PO"},"agent_actions":{"view_html":"https://pith.science/pith/BFIWE5POTHK4TZEPD27QE5DGFN","download_json":"https://pith.science/pith/BFIWE5POTHK4TZEPD27QE5DGFN.json","view_paper":"https://pith.science/paper/BFIWE5PO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.22112&json=true","fetch_graph":"https://pith.science/api/pith-number/BFIWE5POTHK4TZEPD27QE5DGFN/graph.json","fetch_events":"https://pith.science/api/pith-number/BFIWE5POTHK4TZEPD27QE5DGFN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BFIWE5POTHK4TZEPD27QE5DGFN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BFIWE5POTHK4TZEPD27QE5DGFN/action/storage_attestation","attest_author":"https://pith.science/pith/BFIWE5POTHK4TZEPD27QE5DGFN/action/author_attestation","sign_citation":"https://pith.science/pith/BFIWE5POTHK4TZEPD27QE5DGFN/action/citation_signature","submit_replication":"https://pith.science/pith/BFIWE5POTHK4TZEPD27QE5DGFN/action/replication_record"}},"created_at":"2026-07-05T11:11:10.087015+00:00","updated_at":"2026-07-05T11:11:10.087015+00:00"}