让批评者在生成过程中实时修正图像编辑错误，提升精度。
Inline Critic Steers Image Editing
- 引入可学习的中间批评者，实时指导模型修正
- 在GEdit-Bench上达7.89分，比基线高9.4分
- 适合需要高精度图像编辑的开发者与研究者
基于指令的图像编辑在不同区域和案例间难度差异显著，促使人们关注应将修正集中在模型薄弱环节。现有方法的修正信号出现过晚，通常在完整生成或去噪完成后才介入。本文探究是否能在前向传播过程中实时提供反馈。通过对冻结的图像编辑模型进行探测发现，尽管生成能力仅在最后几层显现，但错误模式早在早期层中已确定（与最终层误差图的相关性ρ=0.83）。基于此，提出Inline Critic：一个在中间层对模型预测进行批判并引导隐藏状态修正生成的可学习标记。设计三阶段训练流程以稳定从学会批评到引导生成的过程。实验结果表明，该方法在GEdit-Bench上取得7.89分，相比同主干网络提升9.4分，在RISEBench上实现领先效果，并在KRIS-Bench上以81.92分超越GPT-4o，成为最强开源结果。进一步分析显示，批评者确实在后续层中改变了模型注意力与预测更新。
原文摘要 · Abstract (English)
Instruction-based image editing exhibits heterogeneous difficulty not only across cases but also across regions of an image, motivating refinement approaches that allocate correction to where the model struggles. Existing refinement signals arrive late, after a fully generated image or a completed denoising step. We ask whether such a signal can act within an ongoing forward pass. To investigate this, we probe a frozen image-editing model and find that although generation capability emerges only in the last few layers, the error pattern is already set in early layers (rank correlation \r{ho} = 0.83 with the final-layer error map). Based on this, we introduce Inline Critic, a learnable token that critiques a frozen model's predictions at its intermediate layers and steers its hidden states to refine generation during the forward pass. A three-stage recipe is proposed to stabilize the training from learning how to critique to steering generation. As a result, we achieve state of the art on GEdit-Bench (7.89), a +9.4 gain on RISEBench over the same backbone, and the strongest open-source result on KRIS-Bench (81.92, surpassing GPT-4o). We further provide analyses showing that the critic genuinely shapes the model's attention and prediction updates at subsequent layers.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。