{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:7S5JAHO2JSSE3LL5RJPO6XWLBC","short_pith_number":"pith:7S5JAHO2","schema_version":"1.0","canonical_sha256":"fcba901dda4ca44dad7d8a5eef5ecb08a9ee8ef50afda007280df70e74218923","source":{"kind":"arxiv","id":"2309.04461","version":2},"attestation_state":"computed","paper":{"title":"Measuring and Improving Chain-of-Thought Reasoning in Vision-Language Models","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.CV","cs.LG"],"primary_cat":"cs.CL","authors_text":"Ajay Divakaran, Heng Ji, Karan Sikka, Michael Cogswell, Yangyi Chen","submitted_at":"2023-09-08T17:49:44Z","abstract_excerpt":"Vision-language models (VLMs) have recently demonstrated strong efficacy as visual assistants that can parse natural queries about the visual content and generate human-like outputs. In this work, we explore the ability of these models to demonstrate human-like reasoning based on the perceived information. To address a crucial concern regarding the extent to which their reasoning capabilities are fully consistent and grounded, we also measure the reasoning consistency of these models. We achieve this by proposing a chain-of-thought (CoT) based consistency measure. However, such an evaluation r"},"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":"2309.04461","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-09-08T17:49:44Z","cross_cats_sorted":["cs.CV","cs.LG"],"title_canon_sha256":"f4719900f1bb163b6a647f670f6bd38bc2258dbe7a45e9de221235ee64445997","abstract_canon_sha256":"aab3868ab8313be1a5d90700bd79fefc7b42cb712242b03c0c4bd428a429048a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:58:23.009146Z","signature_b64":"kGqJH++sdPFlBXTnZrG+sMKa6Tgk5+DqZM7aklnZz5dTrJ7xm2lYvHS/3/wtg5OfXiIVBnr5PTCCU6PRdNWEBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fcba901dda4ca44dad7d8a5eef5ecb08a9ee8ef50afda007280df70e74218923","last_reissued_at":"2026-07-05T07:58:23.008629Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:58:23.008629Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Measuring and Improving Chain-of-Thought Reasoning in Vision-Language Models","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.CV","cs.LG"],"primary_cat":"cs.CL","authors_text":"Ajay Divakaran, Heng Ji, Karan Sikka, Michael Cogswell, Yangyi Chen","submitted_at":"2023-09-08T17:49:44Z","abstract_excerpt":"Vision-language models (VLMs) have recently demonstrated strong efficacy as visual assistants that can parse natural queries about the visual content and generate human-like outputs. In this work, we explore the ability of these models to demonstrate human-like reasoning based on the perceived information. To address a crucial concern regarding the extent to which their reasoning capabilities are fully consistent and grounded, we also measure the reasoning consistency of these models. We achieve this by proposing a chain-of-thought (CoT) based consistency measure. However, such an evaluation r"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.04461","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/2309.04461/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":"2309.04461","created_at":"2026-07-05T07:58:23.008696+00:00"},{"alias_kind":"arxiv_version","alias_value":"2309.04461v2","created_at":"2026-07-05T07:58:23.008696+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.04461","created_at":"2026-07-05T07:58:23.008696+00:00"},{"alias_kind":"pith_short_12","alias_value":"7S5JAHO2JSSE","created_at":"2026-07-05T07:58:23.008696+00:00"},{"alias_kind":"pith_short_16","alias_value":"7S5JAHO2JSSE3LL5","created_at":"2026-07-05T07:58:23.008696+00:00"},{"alias_kind":"pith_short_8","alias_value":"7S5JAHO2","created_at":"2026-07-05T07:58:23.008696+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2311.05232","citing_title":"A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questions","ref_index":49,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06422","citing_title":"When to Call an Apple Red: Humans Follow Introspective Rules, VLMs Don't","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.02035","citing_title":"VIDA: A dataset for Visually Dependent Ambiguity in Multimodal Machine Translation","ref_index":30,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7S5JAHO2JSSE3LL5RJPO6XWLBC","json":"https://pith.science/pith/7S5JAHO2JSSE3LL5RJPO6XWLBC.json","graph_json":"https://pith.science/api/pith-number/7S5JAHO2JSSE3LL5RJPO6XWLBC/graph.json","events_json":"https://pith.science/api/pith-number/7S5JAHO2JSSE3LL5RJPO6XWLBC/events.json","paper":"https://pith.science/paper/7S5JAHO2"},"agent_actions":{"view_html":"https://pith.science/pith/7S5JAHO2JSSE3LL5RJPO6XWLBC","download_json":"https://pith.science/pith/7S5JAHO2JSSE3LL5RJPO6XWLBC.json","view_paper":"https://pith.science/paper/7S5JAHO2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2309.04461&json=true","fetch_graph":"https://pith.science/api/pith-number/7S5JAHO2JSSE3LL5RJPO6XWLBC/graph.json","fetch_events":"https://pith.science/api/pith-number/7S5JAHO2JSSE3LL5RJPO6XWLBC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7S5JAHO2JSSE3LL5RJPO6XWLBC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7S5JAHO2JSSE3LL5RJPO6XWLBC/action/storage_attestation","attest_author":"https://pith.science/pith/7S5JAHO2JSSE3LL5RJPO6XWLBC/action/author_attestation","sign_citation":"https://pith.science/pith/7S5JAHO2JSSE3LL5RJPO6XWLBC/action/citation_signature","submit_replication":"https://pith.science/pith/7S5JAHO2JSSE3LL5RJPO6XWLBC/action/replication_record"}},"created_at":"2026-07-05T07:58:23.008696+00:00","updated_at":"2026-07-05T07:58:23.008696+00:00"}