{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:CM43XZS6BBLCJCL32JUPZEC6QA","short_pith_number":"pith:CM43XZS6","schema_version":"1.0","canonical_sha256":"1339bbe65e085624897bd268fc905e8029c4411b2e895f639b1b6afd6effb7c6","source":{"kind":"arxiv","id":"2305.14985","version":3},"attestation_state":"computed","paper":{"title":"IdealGPT: Iteratively Decomposing Vision and Language Reasoning via Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Gengyu Wang, Hammad A. Ayyubi, Haoxuan You, Kai-Wei Chang, Long Chen, Rui Sun, Shih-Fu Chang, Zhecan Wang","submitted_at":"2023-05-24T10:19:57Z","abstract_excerpt":"The field of vision-and-language (VL) understanding has made unprecedented progress with end-to-end large pre-trained VL models (VLMs). However, they still fall short in zero-shot reasoning tasks that require multi-step inferencing. To achieve this goal, previous works resort to a divide-and-conquer pipeline. In this paper, we argue that previous efforts have several inherent shortcomings: 1) They rely on domain-specific sub-question decomposing models. 2) They force models to predict the final answer even if the sub-questions or sub-answers provide insufficient information. We address these l"},"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":"2305.14985","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2023-05-24T10:19:57Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"b7112ce4a41ba1d057d86d38f383b1c09813cd177ca04ecdb16fcc76f91acaaa","abstract_canon_sha256":"7c713254ebd4b8e8dba3d984d03e3508216ae246cc41371557b23cbd75e88b5a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-19T16:12:12.258169Z","signature_b64":"T8S1w3AeaUkiF1F/yyuLwIuyGoahfz9Nu0lwXuN4YhLuR02s+LDNiUby63q+PvJkFAEuY+z97nBL0qIvGTFFAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1339bbe65e085624897bd268fc905e8029c4411b2e895f639b1b6afd6effb7c6","last_reissued_at":"2026-06-19T16:12:12.257733Z","signature_status":"signed_v1","first_computed_at":"2026-06-19T16:12:12.257733Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"IdealGPT: Iteratively Decomposing Vision and Language Reasoning via Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Gengyu Wang, Hammad A. Ayyubi, Haoxuan You, Kai-Wei Chang, Long Chen, Rui Sun, Shih-Fu Chang, Zhecan Wang","submitted_at":"2023-05-24T10:19:57Z","abstract_excerpt":"The field of vision-and-language (VL) understanding has made unprecedented progress with end-to-end large pre-trained VL models (VLMs). However, they still fall short in zero-shot reasoning tasks that require multi-step inferencing. To achieve this goal, previous works resort to a divide-and-conquer pipeline. In this paper, we argue that previous efforts have several inherent shortcomings: 1) They rely on domain-specific sub-question decomposing models. 2) They force models to predict the final answer even if the sub-questions or sub-answers provide insufficient information. We address these l"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.14985","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/2305.14985/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":"2305.14985","created_at":"2026-06-19T16:12:12.257791+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.14985v3","created_at":"2026-06-19T16:12:12.257791+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.14985","created_at":"2026-06-19T16:12:12.257791+00:00"},{"alias_kind":"pith_short_12","alias_value":"CM43XZS6BBLC","created_at":"2026-06-19T16:12:12.257791+00:00"},{"alias_kind":"pith_short_16","alias_value":"CM43XZS6BBLCJCL3","created_at":"2026-06-19T16:12:12.257791+00:00"},{"alias_kind":"pith_short_8","alias_value":"CM43XZS6","created_at":"2026-06-19T16:12:12.257791+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":2,"sample":[{"citing_arxiv_id":"2307.06435","citing_title":"A Comprehensive Overview of Large Language Models","ref_index":296,"is_internal_anchor":true},{"citing_arxiv_id":"2306.13549","citing_title":"A Survey on Multimodal Large Language Models","ref_index":198,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CM43XZS6BBLCJCL32JUPZEC6QA","json":"https://pith.science/pith/CM43XZS6BBLCJCL32JUPZEC6QA.json","graph_json":"https://pith.science/api/pith-number/CM43XZS6BBLCJCL32JUPZEC6QA/graph.json","events_json":"https://pith.science/api/pith-number/CM43XZS6BBLCJCL32JUPZEC6QA/events.json","paper":"https://pith.science/paper/CM43XZS6"},"agent_actions":{"view_html":"https://pith.science/pith/CM43XZS6BBLCJCL32JUPZEC6QA","download_json":"https://pith.science/pith/CM43XZS6BBLCJCL32JUPZEC6QA.json","view_paper":"https://pith.science/paper/CM43XZS6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.14985&json=true","fetch_graph":"https://pith.science/api/pith-number/CM43XZS6BBLCJCL32JUPZEC6QA/graph.json","fetch_events":"https://pith.science/api/pith-number/CM43XZS6BBLCJCL32JUPZEC6QA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CM43XZS6BBLCJCL32JUPZEC6QA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CM43XZS6BBLCJCL32JUPZEC6QA/action/storage_attestation","attest_author":"https://pith.science/pith/CM43XZS6BBLCJCL32JUPZEC6QA/action/author_attestation","sign_citation":"https://pith.science/pith/CM43XZS6BBLCJCL32JUPZEC6QA/action/citation_signature","submit_replication":"https://pith.science/pith/CM43XZS6BBLCJCL32JUPZEC6QA/action/replication_record"}},"created_at":"2026-06-19T16:12:12.257791+00:00","updated_at":"2026-06-19T16:12:12.257791+00:00"}