FP64 is All You Need: Rethinking Failure Modes in Physics-Informed Neural Networks
Physics Informed Neural Networks (PINNs) often exhibit failure modes in which the PDE residual loss converges while the solution error stays large, a phenomenon traditionally blamed on local optima separated from the true solution by steep loss barriers. We challenge this understanding by demonstrate that the real culprit is insufficient arithmetic precision: with standard FP32, the LBFGS optimizer prematurely satisfies its convergence test, freezing the network in a spurious failure phase. Simply upgrading to FP64 rescues optimization, enabling vanilla PINNs to solve PDEs without any failure modes. These results reframe PINN failure modes as precision induced stalls rather than inescapable local minima and expose a three stage training dynamic unconverged, failure, success whose boundaries shift with numerical precision. Our findings emphasize that rigorous arithmetic precision is the key to dependable PDE solving with neural networks.
Code (1)
Tasks
AllSimilar Papers 제목 키워드 기반
Parameterized Physics-informed Neural Networks for Parameterized PDEs
Complex physical systems are often described by partial differential equations (PDEs) that depend on parameters such as the Reynolds number in fluid mechanics. In applications such as design optimization or uncertainty q…
Uncertainty QuantificationA physics-informed neural network for wind turbine main bearing fatigue
Unexpected main bearing failure on a wind turbine causes unwanted maintenance and increased operation costs (mainly due to crane, parts, labor, and production loss). Unfortunately, historical data indicates that failure …
Graph RegressionGraph-to-SequencePhysics-informed machine learningMitigating Propagation Failures in Physics-informed Neural Networks using Retain-Resample-Release (R3) Sampling
Despite the success of physics-informed neural networks (PINNs) in approximating partial differential equations (PDEs), PINNs can sometimes fail to converge to the correct solution in problems involving complicated PDEs.…
Physics-informed neural networks for highly compressible flows
While physics-informed neural networks have been shown to accurately solve a wide range of fluid dynamics problems, their effectivity on highly compressible flows is so far limited. In particular, they struggle with tran…
Physics-informed machine learningSub-Sequential Physics-Informed Learning with State Space Model
Physics-Informed Neural Networks (PINNs) are a kind of deep-learning-based numerical solvers for partial differential equations (PDEs). Existing PINNs often suffer from failure modes of being unable to propagate patterns…