CoT-VLA: Visual Chain-of-Thought Reasoning for Vision-Language-Action Models

By Qingqing Zhao, Yao Lu, Moo Jin Kim, Zipeng Fu, Zhuoyang Zhang, Yecheng Wu, Zhaoshuo Li, Qianli Ma, Song Han, Chelsea Finn and 5 more. Computer Vision and Pattern Recognition, 2025

497 citations recorded by the source corpus.

Vision-language-action models (VLAs) have shown potential in leveraging pretrained vision-language models and diverse robot demonstrations for learning generalizable sensorimotor control. While this paradigm effectively utilizes large-scale data from both robotic and non-robotic sources, current VLAs primarily focus on direct input–output mappings, lacking the intermediate reasoning steps crucial for complex manipulation tasks. As a result, existing VLAs lack temporal planning or reasoning capabilities. In this paper, we introduce a method that incorporates explicit visual chain-of-thought (CoT) reasoning into vision-language-action models (VLAs) by predicting future image frames autoregressively as visual goals before generating a short action sequence to achieve these goals. We introduce CoT-VLA, a state-of-the-art 7B VLA that can understand and generate visual and action tokens. Our experimental results demonstrate that CoT-VLA achieves strong performance, outperforming the state-of-the-art VLA model by 17% in real-world manipulation tasks and 6% in simulation benchmarks. Videos are available at: https://cot-vla.github.io/.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/2d7f3a99e916fc80ff890d109699f9682253e66d

Citations in the corpus

• π0.5: a Vision-Language-Action Model with Open-World Generalization (2025, unknown): “…that augmenting end-to-end policies with high-level reasoning can significantly improve performance for long-horizon tasks [2, 36, 44, 74, 71, 4, 16, 11, 53, 88, 51, 59, 13, 70, 91, 65, 72, 47, 76, 89], particularly when high-level subtask inference can benefit from large pre-trained LLMs and VLMs.”
• V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning (2025, unknown): “Other works have explored the integration of generative modeling into policy learning (Du et al., 2024; Wu et al., 2023a; Zhao et al., 2025; Zhu et al., 2025; Du et al., 2023; Zheng et al., 2025; Rajasegaran et al., 2025).”