改进自回归图像生成的思维链效率
Improving Chain-of-Thought Efficiency for Autoregressive Image Generation
AI总结：
针对自回归图像生成中思维链的冗余问题，提出ShortCoTI轻量优化框架，通过自适应奖励强化学习将提示推理长度减少54%，同时保持或提升图像质量。
摘要：
自回归多模态大语言模型近来因基础模型的进步而在图像生成领域广受欢迎。为增强对齐性和细节，较新的方法采用思维链（CoT）推理，在图像合成之前将用户输入扩展为精细的提示。然而，这种策略可能引入不必要的冗余——我们称之为“视觉过度思考”的现象——这会增加计算成本，并可能引入与原始提示相矛盾的细节。在本工作中，我们探索如何生成更简洁的思维链序列以实现更高效的图像生成。我们提出ShortCoTI，一个轻量级优化框架，在保持输出图像质量的同时鼓励更简洁的思维链。ShortCoTI通过一个根据每个任务估计难度进行缩放的自适应函数来奖励更简洁的提示。将该奖励纳入强化学习范式，可将提示推理长度减少54%，同时在多个基准（T2I-CompBench、GenEval）上保持或略微提升质量指标。定性分析表明，我们的方法消除了冗长的解释和重复的细化，产生既简洁又语义丰富的推理提示。因此，ShortCoTI在不损害生成图像的保真度或视觉吸引力的前提下提高了计算效率。
英文摘要：
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.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
