{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:Q65S5ERWN3ZBMXNTK66SRKACID","short_pith_number":"pith:Q65S5ERW","schema_version":"1.0","canonical_sha256":"87bb2e92366ef2165db357bd28a80240e79b5ac5ec2d4efb435bbb79a45b75e6","source":{"kind":"arxiv","id":"2402.11684","version":2},"attestation_state":"computed","paper":{"title":"ALLaVA: Harnessing GPT4V-Synthesized Data for Lite Vision-Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Benyou Wang, Guiming Hardy Chen, Jianquan Li, Junying Chen, Ruifei Zhang, Shunian Chen, Xiangbo Wu, Xiang Wan, Zhihong Chen, Zhiyi Zhang","submitted_at":"2024-02-18T19:26:49Z","abstract_excerpt":"Large vision-language models (LVLMs) have shown premise in a broad range of vision-language tasks with their strong reasoning and generalization capabilities. However, they require considerable computational resources for training and deployment. This study aims to bridge the performance gap between traditional-scale LVLMs and resource-friendly lite versions by adopting high-quality training data. To this end, we propose a comprehensive pipeline for generating a synthetic dataset. The key idea is to leverage strong proprietary models to generate (i) fine-grained image annotations for vision-la"},"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":"2402.11684","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-02-18T19:26:49Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"a0bc00dfb7156eebdaef3999fd45133da6ea0ec7176ff248e2e45132d0bca1ee","abstract_canon_sha256":"4b22facfa9301417f3da7928f04a5c1d135ad8e0456d5a5e15dd715eeb2be1db"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-23T22:17:25.760227Z","signature_b64":"1JeO/fKtZ8mmHg+t+rdMWjVFuXgrX0KkKeA5WkQDqc/kWIcO1XeIOzI6LA0DSyKGZByIclCWeSHf99ZoJAV9Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"87bb2e92366ef2165db357bd28a80240e79b5ac5ec2d4efb435bbb79a45b75e6","last_reissued_at":"2026-05-23T22:17:25.756870Z","signature_status":"signed_v1","first_computed_at":"2026-05-23T22:17:25.756870Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ALLaVA: Harnessing GPT4V-Synthesized Data for Lite Vision-Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Benyou Wang, Guiming Hardy Chen, Jianquan Li, Junying Chen, Ruifei Zhang, Shunian Chen, Xiangbo Wu, Xiang Wan, Zhihong Chen, Zhiyi Zhang","submitted_at":"2024-02-18T19:26:49Z","abstract_excerpt":"Large vision-language models (LVLMs) have shown premise in a broad range of vision-language tasks with their strong reasoning and generalization capabilities. However, they require considerable computational resources for training and deployment. This study aims to bridge the performance gap between traditional-scale LVLMs and resource-friendly lite versions by adopting high-quality training data. To this end, we propose a comprehensive pipeline for generating a synthetic dataset. The key idea is to leverage strong proprietary models to generate (i) fine-grained image annotations for vision-la"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.11684","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/2402.11684/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":"2402.11684","created_at":"2026-05-23T22:17:25.757042+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.11684v2","created_at":"2026-05-23T22:17:25.757042+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.11684","created_at":"2026-05-23T22:17:25.757042+00:00"},{"alias_kind":"pith_short_12","alias_value":"Q65S5ERWN3ZB","created_at":"2026-05-23T22:17:25.757042+00:00"},{"alias_kind":"pith_short_16","alias_value":"Q65S5ERWN3ZBMXNT","created_at":"2026-05-23T22:17:25.757042+00:00"},{"alias_kind":"pith_short_8","alias_value":"Q65S5ERW","created_at":"2026-05-23T22:17:25.757042+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":23,"internal_anchor_count":23,"sample":[{"citing_arxiv_id":"2412.04468","citing_title":"NVILA: Efficient Frontier Visual Language Models","ref_index":95,"is_internal_anchor":true},{"citing_arxiv_id":"2502.13923","citing_title":"Qwen2.5-VL Technical Report","ref_index":2,"is_internal_anchor":true},{"citing_arxiv_id":"2504.09925","citing_title":"FLARE: Fully Integration of Vision-Language Representations for Deep Cross-Modal Understanding","ref_index":6,"is_internal_anchor":true},{"citing_arxiv_id":"2605.14530","citing_title":"Mitigating Mask Prior Drift and Positional Attention Collapse in Large Diffusion Vision-Language Models","ref_index":4,"is_internal_anchor":true},{"citing_arxiv_id":"2408.04840","citing_title":"mPLUG-Owl3: Towards Long Image-Sequence Understanding in Multi-Modal Large Language Models","ref_index":201,"is_internal_anchor":true},{"citing_arxiv_id":"2605.20033","citing_title":"A Nash Equilibrium Framework For Training-Free Multimodal Step Verification","ref_index":12,"is_internal_anchor":true},{"citing_arxiv_id":"2506.04565","citing_title":"From Standalone LLMs to Integrated Intelligence: A Survey of Compound Al Systems","ref_index":17,"is_internal_anchor":true},{"citing_arxiv_id":"2501.00574","citing_title":"VideoChat-Flash: Hierarchical Compression for Long-Context Video Modeling","ref_index":8,"is_internal_anchor":true},{"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":16,"is_internal_anchor":true},{"citing_arxiv_id":"2412.14164","citing_title":"MetaMorph: Multimodal Understanding and Generation via Instruction Tuning","ref_index":98,"is_internal_anchor":true},{"citing_arxiv_id":"2403.18814","citing_title":"Mini-Gemini: Mining the Potential of Multi-modality Vision Language Models","ref_index":15,"is_internal_anchor":true},{"citing_arxiv_id":"2501.12386","citing_title":"InternVideo2.5: Empowering Video MLLMs with Long and Rich Context Modeling","ref_index":7,"is_internal_anchor":true},{"citing_arxiv_id":"2406.16860","citing_title":"Cambrian-1: A Fully Open, Vision-Centric Exploration of Multimodal LLMs","ref_index":25,"is_internal_anchor":true},{"citing_arxiv_id":"2410.10594","citing_title":"VisRAG: Vision-based Retrieval-augmented Generation on Multi-modality Documents","ref_index":3,"is_internal_anchor":true},{"citing_arxiv_id":"2306.13549","citing_title":"A Survey on Multimodal Large Language Models","ref_index":94,"is_internal_anchor":true},{"citing_arxiv_id":"2605.14530","citing_title":"Mitigating Mask Prior Drift and Positional Attention Collapse in Large Diffusion Vision-Language Models","ref_index":4,"is_internal_anchor":true},{"citing_arxiv_id":"2409.17146","citing_title":"Molmo and PixMo: Open Weights and Open Data for State-of-the-Art Vision-Language Models","ref_index":13,"is_internal_anchor":true},{"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":14,"is_internal_anchor":true},{"citing_arxiv_id":"2605.10641","citing_title":"LLaVA-CKD: Bottom-Up Cascaded Knowledge Distillation for Vision-Language Models","ref_index":18,"is_internal_anchor":true},{"citing_arxiv_id":"2412.10302","citing_title":"DeepSeek-VL2: Mixture-of-Experts Vision-Language Models for Advanced Multimodal Understanding","ref_index":10,"is_internal_anchor":true},{"citing_arxiv_id":"2408.01800","citing_title":"MiniCPM-V: A GPT-4V Level MLLM on Your Phone","ref_index":18,"is_internal_anchor":true},{"citing_arxiv_id":"2408.03326","citing_title":"LLaVA-OneVision: Easy Visual Task Transfer","ref_index":16,"is_internal_anchor":true},{"citing_arxiv_id":"2412.05271","citing_title":"Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling","ref_index":25,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/Q65S5ERWN3ZBMXNTK66SRKACID","json":"https://pith.science/pith/Q65S5ERWN3ZBMXNTK66SRKACID.json","graph_json":"https://pith.science/api/pith-number/Q65S5ERWN3ZBMXNTK66SRKACID/graph.json","events_json":"https://pith.science/api/pith-number/Q65S5ERWN3ZBMXNTK66SRKACID/events.json","paper":"https://pith.science/paper/Q65S5ERW"},"agent_actions":{"view_html":"https://pith.science/pith/Q65S5ERWN3ZBMXNTK66SRKACID","download_json":"https://pith.science/pith/Q65S5ERWN3ZBMXNTK66SRKACID.json","view_paper":"https://pith.science/paper/Q65S5ERW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.11684&json=true","fetch_graph":"https://pith.science/api/pith-number/Q65S5ERWN3ZBMXNTK66SRKACID/graph.json","fetch_events":"https://pith.science/api/pith-number/Q65S5ERWN3ZBMXNTK66SRKACID/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/Q65S5ERWN3ZBMXNTK66SRKACID/action/timestamp_anchor","attest_storage":"https://pith.science/pith/Q65S5ERWN3ZBMXNTK66SRKACID/action/storage_attestation","attest_author":"https://pith.science/pith/Q65S5ERWN3ZBMXNTK66SRKACID/action/author_attestation","sign_citation":"https://pith.science/pith/Q65S5ERWN3ZBMXNTK66SRKACID/action/citation_signature","submit_replication":"https://pith.science/pith/Q65S5ERWN3ZBMXNTK66SRKACID/action/replication_record"}},"created_at":"2026-05-23T22:17:25.757042+00:00","updated_at":"2026-05-23T22:17:25.757042+00:00"}