2025/10/16 by Xiaolin Gu, Gu, Xinyi, Mao, Jiayuan +11
Computer Science · #Automated reasoning #Computation and Language (cs.CL) #Computer Vision and Pattern Recognition (cs.CV) #Domain Adaptation and Few-Shot Learning #FOS: Computer and information sciences #Focus (optics) #Key (lock) #Machine Learning (cs.LG) #Multimodal Machine Learning Applications #Perception #Process (computing) #Set (abstract data type) #Topic Modeling #Training set #Visual reasoning
paper · pdf · doi:10.48550/arxiv.2510.15040
published in arXiv (Cornell University) (Cornell University)
openalex publication_date 2025/10/16 · openalex created_date 2025/10/21 · openalex updated_date 2026/08/08
Pretrained multi-modal large language models (MLLMs) demonstrate strong performance on diverse multimodal tasks, but remain limited in reasoning capabilities for domains where annotations are difficult to collect. In this work, we focus on artificial image domains such as charts, rendered documents, and webpages, which are abundant in practice yet lack large-scale human annotated reasoning datasets. We introduce COGS (COmposition-Grounded data Synthesis), a data-efficient framework for equipping MLLMs with advanced reasoning abilities from a small set of seed questions. The key idea is to decompose each seed question into primitive perception and reasoning factors, which can then be systematically recomposed with new images to generate large collections of synthetic question-answer pairs. Each generated question is paired with subquestions and intermediate answers, enabling reinforcement learning with factor-level process rewards. Experiments on chart reasoning show that COGS substantially improves performance on unseen questions, with the largest gains on reasoning-heavy and compositional questions. Moreover, training with a factor-level mixture of different seed data yields better transfer across multiple datasets, suggesting that COGS induces generalizable capabilities rather than dataset-specific overfitting. We further demonstrate that the framework extends beyond charts to other domains such as webpages.