English
Related papers

Related papers: Dimensional Criticality at Grokking Across MLPs an…

200 papers

A key challenge in building theoretical foundations for deep learning is the complex optimization dynamics of neural networks, resulting from the high-dimensional interactions between the large number of network parameters. Such non-trivial…

Machine Learning · Computer Science 2021-12-07 Mohammad Pezeshki , Amartya Mitra , Yoshua Bengio , Guillaume Lajoie

We investigate the crossover of critical behavior for the dynamic phase transition (DPT) in ferromagnetic thin films using Monte Carlo simulations of the kinetic Ising model, focusing on the scaling behavior of the dynamic order parameter…

Statistical Mechanics · Physics 2025-09-18 Erol Vatansever , Mikel Quintana , Andreas Berger

We study the dependence on the spatial dimensionality of different quantities relevant in the description of the Anderson transition by combining numerical calculations in a $3 \leq d \leq 6$ disordered tight binding model with theoretical…

Disordered Systems and Neural Networks · Physics 2009-11-11 Antonio M. Garcia-Garcia , Emilio Cuevas

Phase transitions are prevalent throughout physics, spanning thermal phenomena like water boiling to magnetic transitions in solids. They encompass cosmological phase transitions in the early universe and the transition into a quark-gluon…

Disordered Systems and Neural Networks · Physics 2025-04-03 Farid Madani , Maxime Denis , Pascal Szriftgiser , Jean Claude Garreau , Adam Rançon , Radu Chicireanu

We study Diffusion Schr\"odinger Bridge (DSB) models in the context of dynamical astrophysical systems, specifically tackling observational inverse prediction tasks within Giant Molecular Clouds (GMCs) for star formation. We introduce the…

Instrumentation and Methods for Astrophysics · Physics 2025-11-13 Ye Zhu , Duo Xu , Zhiwei Deng , Jonathan C. Tan , Olga Russakovsky

The absorbing-state transition in the three-dimensional contact process with and without quenched randomness is investigated by means of Monte-Carlo simulations. In the clean case, a reweighting technique is combined with a careful…

Statistical Mechanics · Physics 2012-12-03 Thomas Vojta

We present a field theoretic renormalization group study for the critical behaviour of a uniformly driven diffusive system with quenched disorder, which is modelled by different kinds of potential barriers between sites. Due to their…

Statistical Mechanics · Physics 2015-06-25 V. Becker , H. K. Janssen

The critical behavior at the frozen/active transition in the Domany-Kinzel stochastic cellular automaton (DKCA) is studied {\it via} a surface growth process in (1+1) dimensions. At criticality, this process presents a kinetic roughening…

Statistical Mechanics · Physics 2015-06-24 A. P. F. Atman , Ronald Dickman , J. G. Moreira

Due to their analytical tractability, random matrix ensembles serve as robust platforms for exploring exotic phenomena in systems that are computationally demanding. Building on a companion letter [arXiv:2312.17481], this paper investigates…

Disordered Systems and Neural Networks · Physics 2024-08-01 Weitao Chen , Olivier Giraud , Jiangbin Gong , Gabriel Lemarié

We consider the particle current in the asymmetric avalanche process on a ring. It is known to exhibit a transition from the intermittent to continuous flow at the critical density of particles. The exact expressions for the first two…

Mathematical Physics · Physics 2022-01-05 Anastasiia A. Trofimova , Alexander M. Povolotsky

Phase transitions from an active into an absorbing, inactive state are generically described by the critical exponents of directed percolation (DP), with upper critical dimension d_c = 4. In the framework of single-species…

Condensed Matter · Physics 2009-10-31 Y. Y. Goldschmidt , H. Hinrichsen , M. Howard , U. C. Täuber

Grokking occurs when a model achieves high training accuracy but generalization to unseen test points happens long after that. This phenomenon was initially observed on a class of algebraic problems, such as learning modular arithmetic…

Machine Learning · Statistics 2026-04-02 Marcel Tomàs Bernal , Neil Rohit Mallinar , Mikhail Belkin

For diffusion processes in dimension $d>1$, the statistics of trajectory observables over the time-window $[0,T]$ can be studied via the Feynman-Kac deformations of the Fokker-Planck generator, that can be interpreted as euclidean…

Statistical Mechanics · Physics 2024-01-22 Cecile Monthus

Diffusion-Limited Aggregation (DLA), the canonical model for non-equilibrium fractal growth, emerges from the simple rule of irreversible attachment by random walkers. Despite four decades of study, a unified computational framework…

Statistical Mechanics · Physics 2026-01-07 Satish Prajapati

Grokking -- the abrupt transition from memorization to generalization after prolonged training -- has been linked to confinement on low-dimensional execution manifolds in modular arithmetic. Whether this mechanism extends beyond arithmetic…

Machine Learning · Computer Science 2026-04-06 Yongzhong Xu

Training modern neural networks often relies on large learning rates, operating at the edge of stability, where the optimization dynamics exhibit oscillatory and chaotic behavior. Empirically, this regime often yields improved…

Machine Learning · Computer Science 2026-04-22 Mario Tuci , Caner Korkmaz , Umut Şimşekli , Tolga Birdal

We study the quantum phase transition in the two-dimensional random Ising model in a transverse field by Monte Carlo simulations. We find results similar to those known analytically in one-dimension: the dynamical exponent is infinite and,…

Disordered Systems and Neural Networks · Physics 2016-08-31 C. Pich , A. P. Young

The crossover behavior of various models exhibiting phase transition to absorbing phase with parity conserving class has been investigated by numerical simulations and cluster mean-field method. In case of models exhibiting Z_2 symmetric…

Statistical Mechanics · Physics 2009-11-13 Geza Odor , Nora Menyhard

Continuous phase transitions are studied in a two dimensional nonequilibrium model with an infinite number of absorbing configurations. Spreading from a localized source is characterized by nonuniversal critical exponents, which vary…

Condensed Matter · Physics 2009-10-28 Ronald Dickman

Recent studies have uncovered intriguing phenomena in deep learning, such as grokking, double descent, and emergent abilities in large language models, which challenge human intuition and are crucial for a deeper understanding of neural…

Machine Learning · Computer Science 2024-02-27 Yufei Huang , Shengding Hu , Xu Han , Zhiyuan Liu , Maosong Sun
‹ Prev 1 3 4 5 6 7 10 Next ›