Improving Chain-of-Thought Efficiency for Autoregressive Image Generation
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912632582176768 |
|---|---|
| author | Gu, Zeqi Georgopoulos, Markos Dai, Xiaoliang Ghazvininejad, Marjan Wang, Chu Juefei-Xu, Felix Li, Kunpeng Shi, Yujun He, Zecheng He, Zijian Zhou, Jiawei Davis, Abe Wang, Jialiang |
| author_facet | Gu, Zeqi Georgopoulos, Markos Dai, Xiaoliang Ghazvininejad, Marjan Wang, Chu Juefei-Xu, Felix Li, Kunpeng Shi, Yujun He, Zecheng He, Zijian Zhou, Jiawei Davis, Abe Wang, Jialiang |
| contents | Autoregressive multimodal large language models have recently gained popularity for image generation, driven by advances in foundation models. To enhance alignment and detail, newer approaches employ chain-of-thought (CoT) reasoning, expanding user inputs into elaborated prompts prior to image synthesis. However, this strategy can introduce unnecessary redundancy -- a phenomenon we call visual overthinking -- which increases computational costs and can introduce details that contradict the original prompt. In this work, we explore how to generate more concise CoT sequences for more efficient image generation. We introduce ShortCoTI, a lightweight optimization framework that encourages more concise CoT while preserving output image quality. ShortCoTI rewards more concise prompts with an adaptive function that scales according to an estimated difficulty for each task. Incorporating this reward into a reinforcement learning paradigm reduces prompt reasoning length by 54% while maintaining or slightly improving quality metrics across multiple benchmarks (T2I-CompBench, GenEval). Qualitative analysis shows that our method eliminates verbose explanations and repetitive refinements, producing reasoning prompts that are both concise and semantically rich. As a result, ShortCoTI improves computational efficiency without compromising the fidelity or visual appeal of generated images. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_05593 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Improving Chain-of-Thought Efficiency for Autoregressive Image Generation Gu, Zeqi Georgopoulos, Markos Dai, Xiaoliang Ghazvininejad, Marjan Wang, Chu Juefei-Xu, Felix Li, Kunpeng Shi, Yujun He, Zecheng He, Zijian Zhou, Jiawei Davis, Abe Wang, Jialiang Computer Vision and Pattern Recognition Artificial Intelligence Computation and Language Autoregressive multimodal large language models have recently gained popularity for image generation, driven by advances in foundation models. To enhance alignment and detail, newer approaches employ chain-of-thought (CoT) reasoning, expanding user inputs into elaborated prompts prior to image synthesis. However, this strategy can introduce unnecessary redundancy -- a phenomenon we call visual overthinking -- which increases computational costs and can introduce details that contradict the original prompt. In this work, we explore how to generate more concise CoT sequences for more efficient image generation. We introduce ShortCoTI, a lightweight optimization framework that encourages more concise CoT while preserving output image quality. ShortCoTI rewards more concise prompts with an adaptive function that scales according to an estimated difficulty for each task. Incorporating this reward into a reinforcement learning paradigm reduces prompt reasoning length by 54% while maintaining or slightly improving quality metrics across multiple benchmarks (T2I-CompBench, GenEval). Qualitative analysis shows that our method eliminates verbose explanations and repetitive refinements, producing reasoning prompts that are both concise and semantically rich. As a result, ShortCoTI improves computational efficiency without compromising the fidelity or visual appeal of generated images. |
| title | Improving Chain-of-Thought Efficiency for Autoregressive Image Generation |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence Computation and Language |
| url | https://arxiv.org/abs/2510.05593 |