Computational Physics

Mikhail Khodak, Min Ki Jung, Brian Wynne et al.

This addresses the problem of slow numerical simulations for researchers in computational science, offering a novel approach that avoids the data-hungry training of neural networks.

Haiyang Yu, Yuchao Lin, Xuan Zhang et al.

This addresses a computational bottleneck in physics, chemistry, and materials science by enabling faster electronic structure calculations with strong generalization.

Sanjeev Raja, Ishan Amin, Fabian Pedregosa et al.

This addresses the critical issue of simulation instability in MLFFs for molecular dynamics, enabling more reliable modeling of longer timescales and better observable estimation, particularly in data-scarce scenarios.

Andreas Burger, Luca Thiede, Nikolaj Rønne et al.

This addresses a bottleneck in computational chemistry for tasks like transition state search and vibrational analysis, offering significant speed and accuracy improvements.

Juan Diego Toscano, Daniel T. Chen, George Em Karniadakis

This addresses a major problem for researchers in SciC and SciML by enabling methodological innovation rather than manual implementation, though it represents a new paradigm rather than an incremental improvement.

Pao-Hsiung Chiu, Jian Cheng Wong, Chin Chun Ooi et al.

This work addresses the practical adoption of PINNs in science and engineering by improving scalability and efficiency, representing a significant leap rather than an incremental step.

Sifan Wang, Shyam Sankaran, Xiantao Fan et al.

This addresses the high computational cost of turbulence simulation for scientists and engineers, offering a new paradigm for continuous modeling.

Sumanth Kumar Boya, Deepak Subramani

This addresses the need for efficient and generalizable PDE solvers in engineering applications, offering a novel approach that reduces data requirements and improves accuracy over existing methods.

Quercus Hernandez, Max Win, Thomas C. O'Connor et al.

This addresses the problem of efficiently modeling complex particle systems for researchers in physics and materials science, offering a novel method that is not purely incremental but builds on existing formalism with new learning strategies.

Sifan Wang, Ananyae Kumar Bhartari, Bowen Li et al.

This addresses optimization challenges in PINNs for scientific computing, with incremental implications for broader multi-task learning.

Minglei Lu, Chensen Lin, Martian Maxey et al.

This work addresses the challenge of bridging scales in fluid dynamics for applications like engineering and materials science, representing a novel direction rather than an incremental improvement.

Kinson Vernet

This work addresses the challenge of contested causal structures in simulations, such as antimicrobial resistance spread, by enabling adaptive modeling, representing a new paradigm rather than an incremental improvement.

Liyao Lyu, Xinyue Yu, Hayden Schaeffer

This provides a method for modeling collective behaviors in biological systems, representing a novel approach to learning measure-dependent interactions from data.

Patrick Cook, Danny Jammooa, Morten Hjorth-Jensen et al.

This presents a novel paradigm for machine learning that could impact scientific computing and general ML by offering an alternative to neuron-based models.

Jiawei Sun, Bin Zhao, Dong Wang et al.

This enables real-time, label-free endoscopic imaging in hard-to-reach areas, representing a strong specific gain rather than a broad breakthrough.

Michael Scherbela, Nicholas Gao, Philipp Grohs et al.

This provides a new gold-standard method for fast and accurate ab-initio calculations, potentially enabling advancements in quantum chemistry, solid-state physics, and material design.

Sifan Wang, Zehao Dou, Tong-Rui Liu et al.

This addresses the problem of generating physically accurate continuous functions for applications in fluid dynamics and solid mechanics, representing a novel method for a known bottleneck rather than an incremental improvement.

Koji Hashimoto, Yuji Hirono, Akiyoshi Sannai

This work offers a foundational physics-based framework for analyzing neural network architectures, potentially benefiting researchers in machine learning and theoretical physics.

Tatsunori Taniai, Ryo Igarashi, Yuta Suzuki et al.

This work addresses a fundamental problem in materials science by enabling more efficient and accurate property prediction from crystal structures, with potential applications in materials design and discovery.

Wassim Tenachi, Rodrigo Ibata, Thibaut L. François et al.

This work addresses the need for automated discovery of common governing laws across multiple datasets in fields like physics and astrophysics, representing a novel extension rather than an incremental improvement.