arXiv:2610.33062v1 Announce Type: new Abstract: Unified large multimodal models (ULMMs) are able to perform visual reasoning through interleaved text-image generation, but generating explicit intermediate images incurs significant computational overhead. While visual latent reasoning addresses this, most existing methods rely on vision-language models (VLMs) incapable of visual generation. We propose Oddin, a novel framework enabling latent visual reasoning in ULMMs. Oddin replaces intermediate images with compact latent tokens processed entirely by the understanding branch, reserving the generation branch for the final output image. We train the model via offline self-distillation with three objectives, final image generation losses, text distillation losses, and cross-entropy losses for
arXiv：2610.33062v1 发布类型：新 摘要：统一的大型多模态模型（ULMMs）能够通过交错进行的文本-图像生成来执行视觉推理，但生成显式的中间图像会带来显著的计算开销。尽管视觉潜在推理可以解决这一问题，但现有方法大多依赖于无法进行视觉生成的视觉-语言模型（VLMs）。我们提出了一种名为Oddin的新框架，能够在ULMMs中实现潜在的视觉推理。Oddin用紧凑的潜在标记取代中间图像，这些标记完全由理解分支处理，而生成分支则专门用于生成最终输出图像。我们通过离线自蒸馏训练该模型，包含三项优化目标：最终图像生成损失、文本蒸馏损失以及交叉熵损失，用于
