面向组合奖励的冲突感知加法引导：流模型中的对抗性生成
Conflict-Aware Additive Guidance for Flow Models under Compositional Rewards
AI总结：
本文提出了一种面向组合奖励的冲突感知加法引导方法，用于在流模型中处理对抗性生成问题，通过动态检测和解决梯度冲突来纠正离曼福德漂移，提升了生成保真度。
摘要：
在推理时间进行引导采样可以无需微调就通过解释生成过程为可控轨迹来驱动最先进的扩散和流模型。这提供了一种简单灵活的方式，将外部约束（如成本函数或预训练验证器）注入受控生成中。然而，现有方法在同时组合多个约束时往往失效，导致偏离真实数据曼福德。在本工作中，我们识别出这种离曼福德漂移的根本原因，并发现近似误差随着梯度不一致程度严重增加。基于这些发现，我们提出了一种轻量且可学习的方法，即冲突感知加法引导（g^car），该方法通过动态检测和解决梯度冲突来主动纠正离曼福德漂移。我们验证了g^car在多样化的领域中的有效性，从合成数据集和图像编辑到生成决策规划与控制。我们的结果表明，g^car有效纠正了离曼福德漂移，在生成保真度方面超越了基线方法，同时使用轻量计算。代码可在https://github.com/yuxuehui/CAR-guidance获取。
英文摘要：
Inference-time guided sampling steers state-of-the-art diffusion and flow models without fine-tuning by interpreting the generation process as a controllable trajectory. This provides a simple and flexible way to inject external constraints (e.g., cost functions or pre-trained verifiers) for controlled generation. However, existing methods often fail when composing multiple constraints simultaneously, which leads to deviations from the true data manifold. In this work, we identify root causes of this off-manifold drift and find that the approximation error scales severely with gradient misalignment. Building on these findings, we propose Conflict-Aware Additive Guidance ($g^\text{car}$), a lightweight and learnable method, which actively rectifies off-manifold drift by dynamically detecting and resolving gradient conflicts. We validate $g^\text{car}$ across diverse domains, ranging from synthetic datasets and image editing to generative decision-making for planning and control. Our results demonstrate that $g^\text{car}$ effectively rectifies off-manifold drift, surpassing baselines in generation fidelity while using light compute. Code is available at https://github.com/yuxuehui/CAR-guidance.
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。
