{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:BGCJSIF4JMU2Z655AXNVVPBIEI","short_pith_number":"pith:BGCJSIF4","schema_version":"1.0","canonical_sha256":"09849920bc4b29acfbbd05db5abc2822219e4a69f066e8b710fb58783447cd81","source":{"kind":"arxiv","id":"2312.06109","version":1},"attestation_state":"computed","paper":{"title":"Vary: Scaling up the Vision Vocabulary for Large Vision-Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chunrui Han, Haoran Wei, Jianjian Sun, Jinrong Yang, Jinyue Chen, Liang Zhao, Lingyu Kong, Xiangyu Zhang, Zheng Ge","submitted_at":"2023-12-11T04:26:17Z","abstract_excerpt":"Modern Large Vision-Language Models (LVLMs) enjoy the same vision vocabulary -- CLIP, which can cover most common vision tasks. However, for some special vision task that needs dense and fine-grained vision perception, e.g., document-level OCR or chart understanding, especially in non-English scenarios, the CLIP-style vocabulary may encounter low efficiency in tokenizing the vision knowledge and even suffer out-of-vocabulary problem. Accordingly, we propose Vary, an efficient and effective method to scale up the vision vocabulary of LVLMs. The procedures of Vary are naturally divided into two "},"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":"2312.06109","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2023-12-11T04:26:17Z","cross_cats_sorted":[],"title_canon_sha256":"9d117eada7210740cec4ae288ad3250362c0afa07500a88cc761a4149734d3a0","abstract_canon_sha256":"fc9f6b8af345c4f9ca5f69a6165176f66bfbc21df77bb71f18163352cb527b41"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:22:38.778209Z","signature_b64":"8daiGQjlqjs4fkiy8QQ22DuE5ABEVMq5vNSgzL1izL9IdUIsEHKOPf304f/30ua7nmglwD3u7NqAanl8Wta9Cg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"09849920bc4b29acfbbd05db5abc2822219e4a69f066e8b710fb58783447cd81","last_reissued_at":"2026-07-05T07:22:38.777714Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:22:38.777714Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Vary: Scaling up the Vision Vocabulary for Large Vision-Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Chunrui Han, Haoran Wei, Jianjian Sun, Jinrong Yang, Jinyue Chen, Liang Zhao, Lingyu Kong, Xiangyu Zhang, Zheng Ge","submitted_at":"2023-12-11T04:26:17Z","abstract_excerpt":"Modern Large Vision-Language Models (LVLMs) enjoy the same vision vocabulary -- CLIP, which can cover most common vision tasks. However, for some special vision task that needs dense and fine-grained vision perception, e.g., document-level OCR or chart understanding, especially in non-English scenarios, the CLIP-style vocabulary may encounter low efficiency in tokenizing the vision knowledge and even suffer out-of-vocabulary problem. Accordingly, we propose Vary, an efficient and effective method to scale up the vision vocabulary of LVLMs. The procedures of Vary are naturally divided into two "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.06109","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/2312.06109/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":"2312.06109","created_at":"2026-07-05T07:22:38.777775+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.06109v1","created_at":"2026-07-05T07:22:38.777775+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.06109","created_at":"2026-07-05T07:22:38.777775+00:00"},{"alias_kind":"pith_short_12","alias_value":"BGCJSIF4JMU2","created_at":"2026-07-05T07:22:38.777775+00:00"},{"alias_kind":"pith_short_16","alias_value":"BGCJSIF4JMU2Z655","created_at":"2026-07-05T07:22:38.777775+00:00"},{"alias_kind":"pith_short_8","alias_value":"BGCJSIF4","created_at":"2026-07-05T07:22:38.777775+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":12,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2410.21169","citing_title":"Document Parsing Unveiled: Techniques, Challenges, and Prospects for Structured Information Extraction","ref_index":254,"is_internal_anchor":false},{"citing_arxiv_id":"2402.03766","citing_title":"MobileVLM V2: Faster and Stronger Baseline for Vision Language Model","ref_index":68,"is_internal_anchor":false},{"citing_arxiv_id":"2409.01704","citing_title":"General OCR Theory: Towards OCR-2.0 via a Unified End-to-end Model","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2511.14998","citing_title":"FinCriticalED: A Visual Benchmark for Financial Fact-Level OCR","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2407.03320","citing_title":"InternLM-XComposer-2.5: A Versatile Large Vision Language Model Supporting Long-Contextual Input and Output","ref_index":152,"is_internal_anchor":false},{"citing_arxiv_id":"2401.16420","citing_title":"InternLM-XComposer2: Mastering Free-form Text-Image Composition and Comprehension in Vision-Language Large Model","ref_index":84,"is_internal_anchor":false},{"citing_arxiv_id":"2512.10362","citing_title":"Visual Funnel: Resolving Contextual Blindness in Multimodal Large Language Models","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2409.18839","citing_title":"MinerU: An Open-Source Solution for Precise Document Content Extraction","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2404.16821","citing_title":"How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites","ref_index":121,"is_internal_anchor":false},{"citing_arxiv_id":"2604.23813","citing_title":"ShredBench: Evaluating the Semantic Reasoning Capabilities of Multimodal LLMs in Document Reconstruction","ref_index":23,"is_internal_anchor":false},{"citing_arxiv_id":"2604.11042","citing_title":"Improving Layout Representation Learning Across Inconsistently Annotated Datasets via Agentic Harmonization","ref_index":45,"is_internal_anchor":false},{"citing_arxiv_id":"2501.13106","citing_title":"VideoLLaMA 3: Frontier Multimodal Foundation Models for Image and Video Understanding","ref_index":75,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BGCJSIF4JMU2Z655AXNVVPBIEI","json":"https://pith.science/pith/BGCJSIF4JMU2Z655AXNVVPBIEI.json","graph_json":"https://pith.science/api/pith-number/BGCJSIF4JMU2Z655AXNVVPBIEI/graph.json","events_json":"https://pith.science/api/pith-number/BGCJSIF4JMU2Z655AXNVVPBIEI/events.json","paper":"https://pith.science/paper/BGCJSIF4"},"agent_actions":{"view_html":"https://pith.science/pith/BGCJSIF4JMU2Z655AXNVVPBIEI","download_json":"https://pith.science/pith/BGCJSIF4JMU2Z655AXNVVPBIEI.json","view_paper":"https://pith.science/paper/BGCJSIF4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.06109&json=true","fetch_graph":"https://pith.science/api/pith-number/BGCJSIF4JMU2Z655AXNVVPBIEI/graph.json","fetch_events":"https://pith.science/api/pith-number/BGCJSIF4JMU2Z655AXNVVPBIEI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BGCJSIF4JMU2Z655AXNVVPBIEI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BGCJSIF4JMU2Z655AXNVVPBIEI/action/storage_attestation","attest_author":"https://pith.science/pith/BGCJSIF4JMU2Z655AXNVVPBIEI/action/author_attestation","sign_citation":"https://pith.science/pith/BGCJSIF4JMU2Z655AXNVVPBIEI/action/citation_signature","submit_replication":"https://pith.science/pith/BGCJSIF4JMU2Z655AXNVVPBIEI/action/replication_record"}},"created_at":"2026-07-05T07:22:38.777775+00:00","updated_at":"2026-07-05T07:22:38.777775+00:00"}