用扩散模型学习设计与评分的联合分布，实现离线优化的泛化能力提升。
Learning Design-Score Manifold to Guide Diffusion Models for Offline Optimization
- 构建设计-评分流形，统一正向预测与反向生成过程。
- 在24种单目标和10种多目标任务中超越现有方法。
- 适合需要高效、无交互优化的材料、药物等复杂系统设计。
复杂系统优化（如新药发现、高性能材料设计）面临底层规则未知且评估成本高昂的挑战。离线优化旨在利用预收集数据集，在不与系统交互的前提下优化设计。然而，传统方法常因超出训练数据范围而表现不佳，导致评分预测不准、设计质量下降。本文提出ManGO，一种基于扩散模型的框架，通过学习设计-评分流形，全面捕捉二者间的耦合关系。不同于将设计与评分空间割裂处理的方法，ManGO统一了前向预测与反向生成，实现对训练数据外情形的泛化。核心在于无导数的条件生成引导机制，结合推理时动态调整的去噪路径优化策略。大量实验表明，ManGO在合成任务、机器人控制、材料设计、DNA序列及真实工程优化等多个领域，优于24种单目标和10种多目标优化方法。
原文摘要 · Abstract (English)
Optimizing complex systems, from discovering therapeutic drugs to designing high-performance materials, remains a fundamental challenge across science and engineering, as the underlying rules are often unknown and costly to evaluate. Offline optimization aims to optimize designs for target scores using pre-collected datasets without system interaction. However, conventional approaches may fail beyond training data, predicting inaccurate scores and generating inferior designs. This paper introduces ManGO, a diffusion-based framework that learns the design-score manifold, capturing the design-score interdependencies holistically. Unlike existing methods that treat design and score spaces in isolation, ManGO unifies forward prediction and backward generation, attaining generalization beyond training data. Key to this is its derivative-free guidance for conditional generation, coupled with adaptive inference-time scaling that dynamically optimizes denoising paths. Extensive evaluations demonstrate that ManGO outperforms 24 single- and 10 multi-objective optimization methods across diverse domains, including synthetic tasks, robot control, material design, DNA sequence, and real-world engineering optimization.
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。