{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/mammoth-vl-eliciting-multimodal-reasoning","title":"MAmmoTH-VL: Eliciting Multimodal Reasoning with Instruction Tuning at Scale","arxiv_id":"2412.05237","date":"2024-12-06","proceeding":null,"authors":["Jarvis Guo","Tuney Zheng","Yuelin Bai","Bo Li","YuBo Wang","King Zhu","Yizhi Li","Graham Neubig","Wenhu Chen","Xiang Yue"],"abstract":"Open-source multimodal large language models (MLLMs) have shown significant potential in a broad range of multimodal tasks. However, their reasoning capabilities remain constrained by existing instruction-tuning datasets, which were predominately repurposed from academic datasets such as VQA, AI2D, and ChartQA. These datasets target simplistic tasks, and only provide phrase-level answers without any intermediate rationales. To address these challenges, we introduce a scalable and cost-effective method to construct a large-scale multimodal instruction-tuning dataset with rich intermediate rationales designed to elicit CoT reasoning. Using only open models, we create a dataset containing 12M instruction-response pairs to cover diverse, reasoning-intensive tasks with detailed and faithful rationales. Experiments demonstrate that training MLLMs on this dataset significantly improves reasoning capabilities, achieving state-of-the-art performance on benchmarks such as MathVerse (+8.1%), MMMU-Pro (+7%), and MuirBench (+13.3%). Additionally, the model demonstrates notable improvements of up to 4% on non-reasoning-based benchmarks. Ablation studies further highlight the importance of key components, such as rewriting and self-filtering, in the dataset construction process.","url_abs":"https://arxiv.org/abs/2412.05237v1","url_pdf":"https://arxiv.org/pdf/2412.05237v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"mammoth-vl-eliciting-multimodal-reasoning","repo_url":"https://github.com/mammoth-vl/mammoth-vl","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"multimodal-reasoning","task_name":"Multimodal Reasoning"},{"task_slug":"visual-question-answering-1","task_name":"Visual Question Answering"},{"task_slug":"visual-question-answering","task_name":"Visual Question Answering (VQA)"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/visual-question-answering-on-mm-vet","task":"Visual Question Answering","dataset":"MM-Vet","model":"MAmmoTH-VL-8B","rank_in_archive_order":31,"of":231,"metrics":{"GPT-4 score":"62.3"},"uses_additional_data":false},{"leaderboard":"/sota/visual-question-answering-on-mm-vet","task":"Visual Question Answering","dataset":"MM-Vet","model":"MAmmoTH-VL-8B (SI)","rank_in_archive_order":37,"of":231,"metrics":{"GPT-4 score":"60.6"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/2412.05237","atlas_url":"https://app.syntology.ai/?focus=2412.05237","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}