922 papers
Computational physics bridges the gap between abstract theory and real-world observation by using powerful computers to solve complex physical problems. This field allows scientists to simulate everything from the collision of subatomic particles to the swirling dynamics of galaxies, offering insights that traditional experiments alone cannot provide.
On Gist.Science, we continuously process every new preprint in this category from arXiv to make these breakthroughs accessible to everyone. Each entry is accompanied by both a clear, plain-language explanation and a detailed technical summary, ensuring that researchers and curious readers alike can grasp the significance of the latest findings without getting lost in dense equations.
Below are the latest papers in computational physics, curated to keep you at the forefront of this rapidly evolving discipline.
This paper introduces a frequency-domain general synthetic iterative scheme (GSIS) that efficiently simulates oscillatory rarefied gas flows by coupling mesoscopic kinetic and macroscopic synthetic equations to achieve super-convergence and asymptotic-preserving properties, making it up to three orders of magnitude faster than conventional methods in near-continuum regimes.
The paper introduces cuGUGA, a high-performance GPU-accelerated operator-direct graphical unitary group approach (GUGA) configuration interaction solver that utilizes constant-time algorithms and custom CUDA kernels to achieve significant speedups over existing CPU and PySCF implementations for small-to-medium active spaces while maintaining high numerical accuracy.
This computational study reveals that asymmetric isosceles triangular dielectric nanostructures exhibit distinct lateral optical force responses, including stable zones and abrupt switching bands driven by Fano-resonance-like interference between discrete eigenmodes and continuum states, thereby offering design guidelines for controlling optical forces through structural geometry.
This paper introduces a "Physics-Informed Uncertainty" paradigm that leverages violations of physical laws as a computationally efficient proxy for predictive uncertainty, significantly improving the success rate and reducing the computational cost of AI-driven inverse design for complex frequency-selective surfaces compared to traditional methods.
This study combines machine learning potentials, density functional theory, and experimental validation to reveal that while ordered nitrogen vacancies in VN precipitates mitigate irradiation damage in ARAFM steels, solute additions like chromium disrupt this ordering and accelerate precipitate dissolution under fusion-relevant conditions.
This paper identifies a previously overlooked, robustly optimal dynamical regime in active matter reservoir computing—located just below a critical damping threshold—that leverages intrinsic multi-stage relaxation to achieve high-performance information processing across varying physical parameters and tasks.
This paper introduces DPA3, a scalable multi-layer graph neural network based on line graph series that adheres to scaling laws and demonstrates superior zero-shot generalization across diverse atomistic systems, establishing it as a highly accurate foundation model for large-scale atomistic applications.
This paper introduces a novel test-time computing framework for PDE foundation models that leverages reward-driven inference-time scaling to enhance prediction accuracy and out-of-distribution robustness, particularly for compressible Euler equations, by utilizing computational resources during inference rather than relying solely on extensive pretraining.
This paper presents a novel framework that reformulates Physics-Informed Neural Networks as differentiable solvers to efficiently learn continuous solution manifolds for steady-state Darcy flow in heterogeneous systems, enabling accurate, mass-conserving solutions and uncertainty quantification through a single training run that integrates data-driven conductivity representations directly into the physics-informed loss function.
This paper presents an ultrafast electrostatic modeling framework for plasmonic nanoparticles of arbitrary geometry that achieves rapid spectral response calculations by projecting the Neumann-Poincaré operator onto a compact dipole basis to avoid large eigenproblems, while incorporating retardation effects via the modified long-wavelength approximation.