通过约束生成路径避免图像模型记忆训练数据，保持画质与提示一致。
Steering Away from Memorization: Reachability-Constrained Reinforcement Learning for Text-to-Image Diffusion
- 将扩散过程视为动态系统，计算可能产生记忆样本的中间状态
- 在推理阶段用最小扰动引导轨迹避开记忆区域，提升多样性与一致性
- 无需修改模型即可插件式使用，适合注重安全生成的场景
文本到图像扩散模型常会记忆训练数据，导致泛化能力不足。现有缓解方法通常以牺牲图像质量或提示对齐为代价。为此，我们提出推理时的可到达性感知扩散引导（RADS），在不改变扩散主干的前提下，通过将去噪过程建模为动力系统，利用可达性分析近似“逆向可达管”——即必然演化为记忆样本的中间状态集合。随后将缓解策略建模为带约束的强化学习问题，让策略在提示嵌入空间中学习最小扰动以引导轨迹远离记忆区域。实证表明，RADS在生成多样性（SSCD）、质量（FID）和对齐度（CLIP）之间的帕累托前沿上优于当前最优基线。关键优势在于无需修改模型，提供即插即用的安全生成方案。
原文摘要 · Abstract (English)
Text-to-image diffusion models often memorize training data, revealing a fundamental failure to generalize beyond the training set. Current mitigation strategies typically sacrifice image quality or prompt alignment to reduce memorization. To address this, we propose Reachability-Aware Diffusion Steering (RADS), an inference-time framework that prevents memorization while preserving generation fidelity. RADS models the diffusion denoising process as a dynamical system and applies concepts from reachability analysis to approximate the "backward reachable tube"--the set of intermediate states that inevitably evolve into memorized samples. We then formulate mitigation as a constrained reinforcement learning (RL) problem, where a policy learns to steer the trajectory away from memorization via minimal perturbations in the caption embedding space. Empirical evaluations show that RADS achieves a superior Pareto frontier between generation diversity (SSCD), quality (FID), and alignment (CLIP) compared to state-of-the-art baselines. Crucially, RADS provides robust mitigation without modifying the diffusion backbone, offering a plug-and-play solution for safe generation. Our website is available at: https://s-karnik.github.io/rads-memorization-project-page/.
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。