{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:3LLQPAT5JLYXNBS4YAVAE622NF","short_pith_number":"pith:3LLQPAT5","schema_version":"1.0","canonical_sha256":"dad707827d4af176865cc02a027b5a6945d09c2cdf5bba35c2b403f4060db6a7","source":{"kind":"arxiv","id":"2403.04732","version":3},"attestation_state":"computed","paper":{"title":"How Far Are We from Intelligent Visual Deductive Reasoning?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.CV"],"primary_cat":"cs.AI","authors_text":"He Bai, Jiatao Gu, Josh Susskind, Navdeep Jaitly, Ruixiang Zhang, Shuangfei Zhai, Yizhe Zhang","submitted_at":"2024-03-07T18:35:54Z","abstract_excerpt":"Vision-Language Models (VLMs) have recently demonstrated incredible strides on diverse vision language tasks. We dig into vision-based deductive reasoning, a more sophisticated but less explored realm, and find previously unexposed blindspots in the current SOTA VLMs. Specifically, we leverage Raven's Progressive Matrices (RPMs), to assess VLMs' abilities to perform multi-hop relational and deductive reasoning relying solely on visual clues. We perform comprehensive evaluations of several popular VLMs employing standard strategies such as in-context learning, self-consistency, and Chain-of-tho"},"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":"2403.04732","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-03-07T18:35:54Z","cross_cats_sorted":["cs.CL","cs.CV"],"title_canon_sha256":"455294704e75cc6940df84e20d77a42b1c5548dd58089d8cc77c4eb5bc9352e6","abstract_canon_sha256":"bf258a8230219837a3ad4c537f41327c887dc2477e528584e08591ce30819919"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:13:50.483452Z","signature_b64":"t3aNRs0BbLvDQgHX1pSvk+sBNOYZ2as3LC/bxINTvGOkv2llweaC9pELqPd2j8219US7gsV+5VMSqO+EqgXYDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dad707827d4af176865cc02a027b5a6945d09c2cdf5bba35c2b403f4060db6a7","last_reissued_at":"2026-07-05T09:13:50.483024Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:13:50.483024Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"How Far Are We from Intelligent Visual Deductive Reasoning?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.CV"],"primary_cat":"cs.AI","authors_text":"He Bai, Jiatao Gu, Josh Susskind, Navdeep Jaitly, Ruixiang Zhang, Shuangfei Zhai, Yizhe Zhang","submitted_at":"2024-03-07T18:35:54Z","abstract_excerpt":"Vision-Language Models (VLMs) have recently demonstrated incredible strides on diverse vision language tasks. We dig into vision-based deductive reasoning, a more sophisticated but less explored realm, and find previously unexposed blindspots in the current SOTA VLMs. Specifically, we leverage Raven's Progressive Matrices (RPMs), to assess VLMs' abilities to perform multi-hop relational and deductive reasoning relying solely on visual clues. We perform comprehensive evaluations of several popular VLMs employing standard strategies such as in-context learning, self-consistency, and Chain-of-tho"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.04732","kind":"arxiv","version":3},"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/2403.04732/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":"2403.04732","created_at":"2026-07-05T09:13:50.483081+00:00"},{"alias_kind":"arxiv_version","alias_value":"2403.04732v3","created_at":"2026-07-05T09:13:50.483081+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.04732","created_at":"2026-07-05T09:13:50.483081+00:00"},{"alias_kind":"pith_short_12","alias_value":"3LLQPAT5JLYX","created_at":"2026-07-05T09:13:50.483081+00:00"},{"alias_kind":"pith_short_16","alias_value":"3LLQPAT5JLYXNBS4","created_at":"2026-07-05T09:13:50.483081+00:00"},{"alias_kind":"pith_short_8","alias_value":"3LLQPAT5","created_at":"2026-07-05T09:13:50.483081+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.23141","citing_title":"VisAnalog: A Diagnostic Suite for Visual Concept Transfer on Natural Images","ref_index":30,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3LLQPAT5JLYXNBS4YAVAE622NF","json":"https://pith.science/pith/3LLQPAT5JLYXNBS4YAVAE622NF.json","graph_json":"https://pith.science/api/pith-number/3LLQPAT5JLYXNBS4YAVAE622NF/graph.json","events_json":"https://pith.science/api/pith-number/3LLQPAT5JLYXNBS4YAVAE622NF/events.json","paper":"https://pith.science/paper/3LLQPAT5"},"agent_actions":{"view_html":"https://pith.science/pith/3LLQPAT5JLYXNBS4YAVAE622NF","download_json":"https://pith.science/pith/3LLQPAT5JLYXNBS4YAVAE622NF.json","view_paper":"https://pith.science/paper/3LLQPAT5","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2403.04732&json=true","fetch_graph":"https://pith.science/api/pith-number/3LLQPAT5JLYXNBS4YAVAE622NF/graph.json","fetch_events":"https://pith.science/api/pith-number/3LLQPAT5JLYXNBS4YAVAE622NF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3LLQPAT5JLYXNBS4YAVAE622NF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3LLQPAT5JLYXNBS4YAVAE622NF/action/storage_attestation","attest_author":"https://pith.science/pith/3LLQPAT5JLYXNBS4YAVAE622NF/action/author_attestation","sign_citation":"https://pith.science/pith/3LLQPAT5JLYXNBS4YAVAE622NF/action/citation_signature","submit_replication":"https://pith.science/pith/3LLQPAT5JLYXNBS4YAVAE622NF/action/replication_record"}},"created_at":"2026-07-05T09:13:50.483081+00:00","updated_at":"2026-07-05T09:13:50.483081+00:00"}