The remarkable success of the autoregressive paradigm has made significant advancement in Multimodal Large Language Models (MLLMs), with powerful models like Show-o, Transfusion and Emu3 achieving notable progress in unified image understanding and generation. For the first time, we uncover a common phenomenon: the understanding capabilities of MLLMs are typically stronger than their generative capabilities, with a significant gap between the two. Building on this insight, we propose HermesFlow, a simple yet general framework designed to seamlessly bridge the gap between understanding and generation in MLLMs. Specifically, we take the homologous data as input to curate homologous preference data of both understanding and generation. Through Pair-DPO and self-play iterative optimization, HermesFlow effectively aligns multimodal understanding and generation using homologous preference data. Extensive experiments demonstrate the significant superiority of our approach over prior methods, particularly in narrowing the gap between multimodal understanding and generation. These findings highlight the potential of HermesFlow as a general alignment framework for next-generation multimodal foundation models. Code: https://github.com/Gen-Verse/HermesFlow

翻译：自回归范式的显著成功极大地推动了多模态大语言模型（MLLMs）的发展，诸如 Show-o、Transfusion 和 Emu3 等强大模型在统一的图像理解与生成方面取得了显著进展。我们首次揭示了一个普遍现象：MLLMs 的理解能力通常强于其生成能力，两者之间存在显著差距。基于这一洞见，我们提出了 HermesFlow，一个简单而通用的框架，旨在无缝弥合 MLLMs 中理解与生成之间的鸿沟。具体而言，我们以同源数据作为输入，精心构建了涵盖理解与生成的同源偏好数据。通过 Pair-DPO 与自博弈迭代优化，HermesFlow 利用同源偏好数据有效地对齐了多模态理解与生成。大量实验证明，我们的方法相较于先前方法具有显著优越性，特别是在缩小多模态理解与生成之间的差距方面。这些发现凸显了 HermesFlow 作为下一代多模态基础模型的通用对齐框架的潜力。代码：https://github.com/Gen-Verse/HermesFlow