论文标题
可扩展的贝叶斯不确定性定量神经网络电位：承诺和陷阱
Scalable Bayesian Uncertainty Quantification for Neural Network Potentials: Promise and Pitfalls
论文作者
论文摘要
神经网络（NN）电位有望在经典MD力场的计算复杂性中高度准确的分子动力学（MD）模拟。但是，当在其训练领域外应用时，NN的潜在预测可能不准确，从而增加了对不确定性定量的需求（UQ）。贝叶斯建模为UQ提供了数学框架，但是基于Markov Chain Monte Carlo（MCMC）的古典贝叶斯方法在计算上对于NN电位上的计算很难。通过训练图NN的液态水和丙氨酸二肽系统的电势，我们在这里证明了可扩展的贝叶斯uq通过随机梯度MCMC（SG-MCMC）得出可观察到的MD可观察性的可靠不确定性估计。我们表明，冷的后者可以减少所需的训练数据大小，并且对于可靠的UQ，需要多个马尔可夫链。此外，我们发现SG-MCMC和Deep Ensemble方法取得了可比的结果，尽管训练较短，而后者进行了较少的高参数调整。我们表明，这两种方法均可可靠地捕获息肉和认知的不确定性，但不能通过适当的建模来最大程度地降低系统的不确定性，以获得MD可观察物的准确可靠间隔。我们的结果代表了迈向准确的UQ的一步，这对于可信赖的NN潜在潜在的MD模拟至关重要。
Neural network (NN) potentials promise highly accurate molecular dynamics (MD) simulations within the computational complexity of classical MD force fields. However, when applied outside their training domain, NN potential predictions can be inaccurate, increasing the need for Uncertainty Quantification (UQ). Bayesian modeling provides the mathematical framework for UQ, but classical Bayesian methods based on Markov chain Monte Carlo (MCMC) are computationally intractable for NN potentials. By training graph NN potentials for coarse-grained systems of liquid water and alanine dipeptide, we demonstrate here that scalable Bayesian UQ via stochastic gradient MCMC (SG-MCMC) yields reliable uncertainty estimates for MD observables. We show that cold posteriors can reduce the required training data size and that for reliable UQ, multiple Markov chains are needed. Additionally, we find that SG-MCMC and the Deep Ensemble method achieve comparable results, despite shorter training and less hyperparameter tuning of the latter. We show that both methods can capture aleatoric and epistemic uncertainty reliably, but not systematic uncertainty, which needs to be minimized by adequate modeling to obtain accurate credible intervals for MD observables. Our results represent a step towards accurate UQ that is of vital importance for trustworthy NN potential-based MD simulations required for decision-making in practice.
