{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:M6RZEZAZMI2FOZ7D2WCSBRREDM","short_pith_number":"pith:M6RZEZAZ","schema_version":"1.0","canonical_sha256":"67a392641962345767e3d58520c6241b0159dafd92efce586feff664345b0512","source":{"kind":"arxiv","id":"2502.14846","version":2},"attestation_state":"computed","paper":{"title":"Scaling Text-Rich Image Understanding via Code-Guided Synthetic Multimodal Data Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Ajay Patel, Andrew Head, Aniruddha Kembhavi, Chris Callison-Burch, Christopher Clark, Luca Weihs, Mark Yatskar, Matt Deitke, Ranjay Krishna, Tanmay Gupta, Yue Yang","submitted_at":"2025-02-20T18:55:30Z","abstract_excerpt":"Reasoning about images with rich text, such as charts and documents, is a critical application of vision-language models (VLMs). However, VLMs often struggle in these domains due to the scarcity of diverse text-rich vision-language data. To address this challenge, we present CoSyn, a framework that leverages the coding capabilities of text-only large language models (LLMs) to automatically create synthetic text-rich multimodal data. Given input text describing a target domain (e.g., \"nutrition fact labels\"), CoSyn prompts an LLM to generate code (Python, HTML, LaTeX, etc.) for rendering synthe"},"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":"2502.14846","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-02-20T18:55:30Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"11fb8b49b6334760245526a1574f127c6faebc2c92704932127d419f7dc0fa3b","abstract_canon_sha256":"09293db9f57895cfc3d92cb90e6ed229f77e3ab4d4e100ea1f4890623ce7f2f2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:06:28.666649Z","signature_b64":"ooh0F4zmoQTDXUdsrV1BBQJaoMAkVQQGptdxjTui5GTatS5N2eZWSsDfLknf9IIMjfcTqDwAO+1V113n5178CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"67a392641962345767e3d58520c6241b0159dafd92efce586feff664345b0512","last_reissued_at":"2026-07-05T11:06:28.666129Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:06:28.666129Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Scaling Text-Rich Image Understanding via Code-Guided Synthetic Multimodal Data Generation","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Ajay Patel, Andrew Head, Aniruddha Kembhavi, Chris Callison-Burch, Christopher Clark, Luca Weihs, Mark Yatskar, Matt Deitke, Ranjay Krishna, Tanmay Gupta, Yue Yang","submitted_at":"2025-02-20T18:55:30Z","abstract_excerpt":"Reasoning about images with rich text, such as charts and documents, is a critical application of vision-language models (VLMs). However, VLMs often struggle in these domains due to the scarcity of diverse text-rich vision-language data. To address this challenge, we present CoSyn, a framework that leverages the coding capabilities of text-only large language models (LLMs) to automatically create synthetic text-rich multimodal data. Given input text describing a target domain (e.g., \"nutrition fact labels\"), CoSyn prompts an LLM to generate code (Python, HTML, LaTeX, etc.) for rendering synthe"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.14846","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/2502.14846/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":"2502.14846","created_at":"2026-07-05T11:06:28.666191+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.14846v2","created_at":"2026-07-05T11:06:28.666191+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.14846","created_at":"2026-07-05T11:06:28.666191+00:00"},{"alias_kind":"pith_short_12","alias_value":"M6RZEZAZMI2F","created_at":"2026-07-05T11:06:28.666191+00:00"},{"alias_kind":"pith_short_16","alias_value":"M6RZEZAZMI2FOZ7D","created_at":"2026-07-05T11:06:28.666191+00:00"},{"alias_kind":"pith_short_8","alias_value":"M6RZEZAZ","created_at":"2026-07-05T11:06:28.666191+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":8,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2604.18347","citing_title":"Multilingual Training and Evaluation Resources for Vision-Language Models","ref_index":42,"is_internal_anchor":true},{"citing_arxiv_id":"2606.18216","citing_title":"Zone of Proximal Policy Optimization: Teacher in Prompts, Not Gradients","ref_index":102,"is_internal_anchor":false},{"citing_arxiv_id":"2607.02512","citing_title":"Program-as-Weights: A Programming Paradigm for Fuzzy Functions","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2601.13606","citing_title":"ChartVerse: Scaling Chart Reasoning via Reliable Programmatic Synthesis from Scratch","ref_index":44,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11405","citing_title":"20/20 Vision Language Models: A Prescription for Better VLMs through Data Curation Alone","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11405","citing_title":"20/20 Vision Language Models: A Prescription for Better VLMs through Data Curation Alone","ref_index":38,"is_internal_anchor":false},{"citing_arxiv_id":"2605.03547","citing_title":"Erase Persona, Forget Lore: Benchmarking Multimodal Copyright Unlearning in Large Vision Language Models","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2604.18347","citing_title":"Multilingual Training and Evaluation Resources for Vision-Language Models","ref_index":42,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/M6RZEZAZMI2FOZ7D2WCSBRREDM","json":"https://pith.science/pith/M6RZEZAZMI2FOZ7D2WCSBRREDM.json","graph_json":"https://pith.science/api/pith-number/M6RZEZAZMI2FOZ7D2WCSBRREDM/graph.json","events_json":"https://pith.science/api/pith-number/M6RZEZAZMI2FOZ7D2WCSBRREDM/events.json","paper":"https://pith.science/paper/M6RZEZAZ"},"agent_actions":{"view_html":"https://pith.science/pith/M6RZEZAZMI2FOZ7D2WCSBRREDM","download_json":"https://pith.science/pith/M6RZEZAZMI2FOZ7D2WCSBRREDM.json","view_paper":"https://pith.science/paper/M6RZEZAZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.14846&json=true","fetch_graph":"https://pith.science/api/pith-number/M6RZEZAZMI2FOZ7D2WCSBRREDM/graph.json","fetch_events":"https://pith.science/api/pith-number/M6RZEZAZMI2FOZ7D2WCSBRREDM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/M6RZEZAZMI2FOZ7D2WCSBRREDM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/M6RZEZAZMI2FOZ7D2WCSBRREDM/action/storage_attestation","attest_author":"https://pith.science/pith/M6RZEZAZMI2FOZ7D2WCSBRREDM/action/author_attestation","sign_citation":"https://pith.science/pith/M6RZEZAZMI2FOZ7D2WCSBRREDM/action/citation_signature","submit_replication":"https://pith.science/pith/M6RZEZAZMI2FOZ7D2WCSBRREDM/action/replication_record"}},"created_at":"2026-07-05T11:06:28.666191+00:00","updated_at":"2026-07-05T11:06:28.666191+00:00"}