arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2603.01949cs.LGcs.AIcs.CE
收藏

学习模拟器的概率性改造

Probabilistic Retrofitting of Learned Simulators

The Polymathic AI Collaboration(多智AI协作体);University of Cambridge(剑桥大学);Alan Turing Institute(艾伦·图灵研究所)

Cristiana Diaconu, Miles Cranmer, Richard E. Turner, Tanya Marwah, Payel Mukhopadhyay

AI总结:

提出一种训练高效的策略,通过连续排序概率得分(CRPS)将预训练的确定性模型改造为概率模型,无需从头训练,适用于多种架构,在多个动力系统上显著提升预测性能。

摘要:

建模偏微分方程(PDE)的主流方法依赖于确定性预测,然而许多感兴趣的物理系统本质上是混沌且不确定的。虽然从头训练概率模型是可能的,但计算成本高昂,且无法利用已投入高性能确定性骨干网络的大量资源。在这项工作中,我们采用了一种训练高效的策略,通过使用适当的评分规则——连续排序概率得分(CRPS)进行改造,将预训练的确定性模型转化为概率模型。关键的是,这种方法与架构无关:它通过最小的代码修改,将相同的适应机制应用于不同的模型骨干。该方法在不同规模的预训练中均表现出高效性:对于在单一动力系统上训练的模型,与计算匹配的确定性微调相比,我们实现了滚动CRPS降低20-54%,方差归一化RMSE(VRMSE)提升高达30%。我们进一步在PDE基础模型上验证了该方法,该模型在多个系统上训练并在感兴趣的数据集上改造,结果表明,与确定性微调相比,我们的概率适应在CRPS上提升高达40%,在VRMSE上提升高达15%。通过在不同架构和动力学上的验证,我们的结果表明,概率PDE建模无需从头重新训练,而是可以通过适度的额外训练成本从现有的确定性骨干网络中解锁。

英文摘要:

Dominant approaches for modelling Partial Differential Equations (PDEs) rely on deterministic predictions, yet many physical systems of interest are inherently chaotic and uncertain. While training probabilistic models from scratch is possible, it is computationally expensive and fails to leverage the significant resources already invested in high-performing deterministic backbones. In this work, we adopt a training-efficient strategy to transform pre-trained deterministic models into probabilistic ones via retrofitting with a proper scoring rule: the Continuous Ranked Probability Score (CRPS). Crucially, this approach is architecture-agnostic: it applies the same adaptation mechanism across distinct model backbones with minimal code modifications. The method proves highly effective across different scales of pre-training: for models trained on single dynamical systems, we achieve 20-54% reductions in rollout CRPS and up to 30% improvements in variance-normalised RMSE (VRMSE) relative to compute-matched deterministic fine-tuning. We further validate our approach on a PDE foundation model, trained on multiple systems and retrofitted on the dataset of interest, to show that our probabilistic adaptation yields an improvement of up to 40% in CRPS and up to 15% in VRMSE compared to deterministic fine-tuning. Validated across diverse architectures and dynamics, our results show that probabilistic PDE modelling need not require retraining from scratch, but can be unlocked from existing deterministic backbones with modest additional training cost.

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。

↑