- Rating
- 1
- Confidence
- 3
- Informativeness
- 2
- Presentation
- 2
- Soundness
- 1
Read review
Paper summary
This paper presents underPINN, a modular JAX-based framework for physics-informed machine learning that integrates JIT-compiled training, multiple neural architectures, adaptive collocation strategies, and numerical optimization methods. Through comparisons with PyTorch, NVIDIA PhysicsNeMo, jinns, and DeepXDE, the authors demonstrate significant computational speedups while identifying hidden configuration differences and numerical nondeterminism that can undermine the reliability of cross-framework evaluations.
Critical strengths
-
The paper addresses a relevant challenge in physics-informed machine learning: developing an efficient, extensible PINN framework while ensuring reliable and reproducible cross-framework comparisons.
-
underPINN separates physical models, domain formulations, neural network architectures, optimization, and execution, making it easier to integrate different PDE problems and training strategies.
Critical weaknesses
-
The core methodological novelty is limited. Most of the framework's major components, including JAX compilation, Fourier features, gated networks, FBPINN decomposition, adaptive collocation, and Gauss–Newton optimization, have already been studied extensively. The main contribution is their software integration and systematic evaluation rather than a fundamentally new learning algorithm.
-
Higher computational throughput does not translate into better predictive accuracy. In Table 2, underPINN achieves the lowest mean prediction error on none of the five problems with comparable accuracy results across the evaluated JAX frameworks. This weakens the practical advantage of the proposed framework beyond training speed.
-
The framework's native architecture support remains relatively limited, focusing primarily on conventional MLPs and their gated or Fourier-feature variants. Several influential developments in physics-informed learning, including PirateNets, PINNsFormer, PINNMamba, physics-informed Kolmogorov–Arnold networks (PIKANs), and separable PINNs (SPINNs), are not supported or evaluated. This limitation raises concerns about the framework's extensibility and its applicability to modern PINN research, particularly because some of these architectures require substantial changes to the computation or training pipeline rather than simply replacing the network backbone.
Questions for authors
-
Could the authors demonstrate the integration of at least one non-MLP architecture without modifying the core training engine, and evaluate whether its computational efficiency and numerical accuracy are preserved?
-
The PhysicsNeMo experiments disable its default JIT compilation due to hardware incompatibility, whereas underPINN benefits from JAX compilation. How representative are the reported speedups of a comparison between fully optimized implementations? Could the authors provide additional experiments on hardware where PhysicsNeMo's native compiled execution is supported?
Thank you for the detailed review. We answer the soundness points first, then extensibility (Q1) with new experiments.
W1. Limited methodological novelty. We agree the building blocks are established, and the revision no longer presents their integration as the contribution. What is new (General Response §1):
torch.compilecannot capture the PINN double-backward.W2. underPINN is not the most accurate in Table 2. On matched problems all frameworks run the same algorithm, so accuracy gaps come from configuration. The paper traces each one: jinns' initializer (about half its Steady Heat advantage, A.10), DeepXDE's fixed collocation set, and our default learning rate on ODE Harmonic (Sec. 3.9). A framework that "won" at a matched configuration would signal a mismatch. The revision adds a same-initializer arm to Table 2. Time-to-accuracy is in our response to Reviewer 1.
Q2. PhysicsNeMo without its default JIT. A fair point. On GB10, nvFuser and TorchScript fail to compile; on A100 all paths work. Burgers, 3 seeds, ms/epoch: underPINN 1.8, eager PyTorch 5.78, TorchScript 7.69,
torch.compile(Inductor) 12.82.torch.compilecannot capture the double-backward, so it needs the slowertorch.funcform. PhysicsNeMo with its own defaults (TorchScript activation, CUDA graphs; torch 2.5 has no nvFuser) takes 3.10. A100 becomes the primary comparison.W3 / Q1. Non-MLP architectures without touching the training engine. Done. We added PirateNets, PINNsFormer, PINNMamba and Chebyshev and B-spline PIKANs as network modules, selected by
network.type. No solver, loss or training-loop code changed. Data parallelism, restart and diagnostics work unchanged.Parameter-matched (~130k), 3 seeds, best of two learning rates, 20k Adam epochs; median rel. L2:
† Without its least-squares output initialization, which hurt transport problems. ‡ Two of three seeds.
PirateNet and PIKANs cost 1.2–2× the MLP per epoch. The sequence models cost 5–30×, from their architecture, not our pipeline.
A cautionary result. At equal wall-clock, with a standard recipe (causal weighting, gradient-norm balancing, L-BFGS finish) for all models, most gains vanish. PirateNet's 4.9× Allen–Cahn advantage becomes 0.9×, and a plain MLP solves convection (7.7e-3). Only the gated models keep a clear gain on wave (2.4–4×).
SPINN. It needs a different pipeline (per-axis forward-mode derivatives on factorized grids), so we implemented it separately and tested it against autodiff. It is a different network: 3×64 per axis, 64 separable terms. On Helmholtz it takes 0.47 ms/epoch and reaches 2% error in 1.5 s, about 4× sooner than the fastest framework. On Burgers it does not reach 1% error in 20,000 epochs, with fixed or per-step resampled points (final 6–16%).
Presentation. The main text now reports final results; revision history moves to an appendix.