# Watch Wider and Think Deeper: Collaborative Cross-modal Chain-of-Thought for Complex Visual Reasoning

Source: [Watch Wider and Think Deeper: Collaborative Cross-modal Chain-of-Thought for Complex Visual Reasoning](<https://www.sota2.com/research/paper/695de85d53fdd5c361743977>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-01-08T05:00:18.138Z

Authors: Wenting Lu, Didi Zhu, Tao Shen, Donglin Zhu, Ayong Ye, Chao Wu

Published: 2026-01-04T02:50:55.000Z

## Abstract

Multi-modal reasoning requires the seamless integration of visual and linguistic cues, yet existing Chain-of-Thought methods suffer from two critical limitations in cross-modal scenarios: (1) over-reliance on single coarse-grained image regions, and (2) semantic fragmentation between successive reasoning steps. To address these issues, we propose the CoCoT (Collaborative Coross-modal Thought) framework, built upon two key innovations: a) Dynamic Multi-Region Grounding to adaptively detect the most relevant image regions based on the question, and b) Relation-Aware Reasoning to enable multi-region collaboration by iteratively aligning visual cues to form a coherent and logical chain of thought. Through this approach, we construct the CoCoT-70K dataset, comprising 74,691 high-quality samples with multi-region annotations and structured reasoning chains. Extensive experiments demonstrate that CoCoT significantly enhances complex visual reasoning, achieving an average accuracy improvement of 15.4% on LLaVA-1.5 and 4.0% on Qwen2-VL across six challenging benchmarks. The data and code are available at: https://github.com/deer-echo/CoCoT.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Visual Question Answering on TextVQA (Accuracy Single/Multi)](<https://www.sota2.com/research/sota/visual-question-answering-on-textvqa-accuracy-single-multi.md>): Visual Question Answering; TextVQA; Accuracy (Single): 75.9.

- [Visual Question Answering on InfoVQA](<https://www.sota2.com/research/sota/visual-question-answering-on-infovqa.md>): Visual Question Answering; InfoVQA; Accuracy: not recorded.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2601.02422>)

[PDF](<https://arxiv.org/pdf/2601.02422.pdf>)