917 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 novel, verified algorithm that leverages spectral graph theory and vertex curvatures to efficiently and accurately determine graph isomorphism in polynomial time, successfully resolving challenging instances that defeat classical spectral techniques.
This paper presents a comprehensive linear and numerical analysis of the two-dimensional Kelvin-Helmholtz instability in collisionless plasmas with anisotropic pressure, revealing that the magnetohydrodynamic limit yields significantly larger growth rates, current densities, and magnetic island formation compared to the anisotropic CGL regime where energy is diverted into pressure anisotropies.
Neural-POD is a plug-and-play neural operator framework that learns resolution-invariant, nonlinear orthogonal basis functions directly in function space to overcome discretization limitations in AI4Science models, thereby enhancing generalization and interpretability for complex systems like the Burgers' and Navier-Stokes equations.
This study employs machine learning potentials and molecular dynamics simulations to elucidate the deformation mechanisms and compressive response of NbTaTiZr refractory multi-principal element alloys, revealing significant anisotropy in yield strength and twinning behavior across crystal orientations, a strain-rate-dependent transition from dislocation slip to structural disordering, and the compositional influence of Nb/Ta versus Ti/Zr on mechanical performance.
This paper presents a robust inverse reconstruction framework that integrates scattering theory, Lorentz oscillator modeling, and nonlinear effective medium approximations to accurately retrieve the broadband complex permittivity, constituent composition, and microstructural topology of heterogeneous optical composites from a single infrared extinction spectrum, thereby overcoming the limitations of conventional linear unmixing methods.
This paper presents an exact analytical framework for separating center-of-mass and relative motions in anisotropic two-dimensional magnetoexcitons, revealing new anisotropy-dependent couplings and providing precise, non-perturbative solutions for magnetoexciton properties in materials like monolayer black phosphorus and titanium trisulfide without relying on stationary-center-of-mass approximations.
This paper presents a web-based interactive platform developed at St. Xavier's College, Kolkata, which successfully replicates physical physics laboratory setups to enhance undergraduate students' conceptual understanding and experimental confidence, as evidenced by overwhelmingly positive feedback from users.
This paper introduces Geometric Autoencoders for Bayesian Inversion (GABI), a framework that learns geometry-aware generative models from large datasets of varying physical systems to serve as informative priors for robust, well-calibrated uncertainty quantification in ill-posed inverse problems without requiring knowledge of governing equations.
This study validates the accuracy of the UMA universal machine learning interatomic potential for 25 different MoS2 dopants and demonstrates its ability to efficiently simulate complex phenomena like dopant clustering and layer fracturing, thereby enabling high-throughput computational design of doped MoS2 for targeted applications.
The paper introduces MDIntrinsicDimension, an open-source Python package that estimates the intrinsic dimension of molecular dynamics trajectories using invariant projections and state-of-the-art estimators to provide detailed, time-resolved insights into the flexibility and heterogeneity of macromolecular motions.