用物理模拟器做干预，解决科学领域潜变量干扰下的因果发现难题。
Physical Simulators as Do-Operators: Causal Discovery under Latent Confounders for AI-for-Science
- 将物理模拟器作为真实干预工具，结合因果流匹配模型处理潜变量。
- 仅需约d次单变量干预即可识别d个变量的因果结构，效率接近理论下限。
- 在分子毒性与电池电解质优化中减少57%-58%预测偏差，实用性强。
现有干预式因果发现方法（如IGSP、DCDI、ENCO）假设无潜变量干扰，依赖合成模拟器中的虚拟干预。但在分子设计、材料科学等AI for Science场景中，潜变量普遍存在，而真实干预（如基于物理的模拟）每个数据点需数小时至数日。本文提出CFM-SD（Causal Flow Matching with Simulation Data），利用第一性原理物理模拟器作为皮尔逊干预演算中的do-操作符，同时应对潜变量与真实干预数据。理论上，在物理可实现约束下，d个变量的因果结构可通过O(d)次单变量干预唯一确定。在合成数据上的内在评估中（γ=0.2–0.8），CFM-SD平均F1达0.800，显著优于各基线（F1=0.127–0.562）。在外在评估中，于真实科学数据上实现分子毒性预测与电池电解质优化中57%–58%的偏差降低，验证其超越合成基准的实际价值。
原文摘要 · Abstract (English)
Existing interventional causal discovery methods -- IGSP, DCDI, ENCO -- assume causal sufficiency (no latent confounders) and rely on virtual interventions in synthetic simulators. In AI-for-Science settings such as molecular design and materials science, latent confounders are ubiquitous and real interventions (e.g., physics-based simulations) require hours to days per data point. We propose CFM-SD (Causal Flow Matching with Simulation Data), which uses first-principles physical simulators as do-operators in Pearl's interventional calculus to simultaneously handle latent confounders and real interventional data. Theoretically, $d$-variable causal structure is identifiable with $O(d)$ single-variable interventions -- the minimum under physical realizability constraints. In Intrinsic Evaluation on synthetic data ($γ=0.2$--$0.8$), CFM-SD achieves average F1$=0.800$ vs. F1$=0.127$--$0.562$ for all baselines. In Extrinsic Evaluation on real scientific data, CFM-SD achieves 57--58\% bias reduction in molecular toxicity prediction and battery electrolyte optimization, demonstrating practical value beyond synthetic benchmarks.
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。