VisReason: A Large-Scale Dataset for Visual Chain-of-Thought Reasoning

TL;DR

This paper introduces VisReason, a large-scale dataset with annotated visual reasoning examples, to enhance multimodal large language models' ability to perform complex, interpretable, and spatially grounded visual reasoning tasks.

Contribution

It presents VisReason and VisReason-Pro datasets, enabling systematic training of MLLMs for human-like visual reasoning with improved accuracy and interpretability.

Findings

Fine-tuning Qwen2.5-VL on VisReason improves reasoning accuracy.

VisReason enhances interpretability and generalization in visual reasoning.

The datasets support multi-domain, step-by-step visual reasoning.

Abstract

Chain-of-Thought (CoT) prompting has proven remarkably effective for eliciting complex reasoning in large language models (LLMs). Yet, its potential in multimodal large language models (MLLMs) remains largely untapped, hindered by the absence of large-scale datasets that capture the rich, spatially grounded reasoning intrinsic to visual understanding. Existing visual-CoT resources are typically small, domain-specific, or lack the human-like stepwise structure necessary for compositional visual reasoning. In this paper, we introduce VisReason, a large-scale dataset designed to advance visual Chain-of-Thought reasoning. VisReason comprises 489K annotated examples spanning four diverse domains, each featuring multi-round, human-like rationales that guide MLLMs through interpretable visual reasoning steps. Building upon this, we curate VisReason-Pro, a 165K subset produced with a stronger expert-level GPT annotator, enriched with detailed reasoning traces and 3D spatial grounding via depth-informed annotations. Fine-tuning the state-of-the-art Qwen2.5-VL model on VisReason and VisReason-Pro yields substantial improvements in step-by-step visual reasoning accuracy, interpretability, and cross-benchmark generalization. These results demonstrate that VisReason equips MLLMs with more systematic and generalizable reasoning capabilities. We envision VisReason as a cornerstone for cultivating human-like visual reasoning, paving the way toward the next generation of multimodal intelligence.