arXiv subjects

Jack T. Parley

Publications and source records attributed to Jack T. Parley.

7 recordsLinked to original sources

arXiv2026

Deep networks learn to parse uniform-depth context-free languages from local statistics

Understanding how the structure of language can be learned from sentences alone is a central question in both cognitive science and machine learning. Studies of the internal representations of Large Language Models (LLMs) support their ability to parse text when predicting the next word, while representing semantic notions independently of surface form. Yet, which data statistics make these feats possible, and how much data is required, remain largely unknown. Probabilistic context-free grammars (PCFGs) provide a tractable testbed for studying these questions. However, prior work has focused either on the post-hoc characterization of the parsing-like algorithms used by trained networks; or on the learnability of PCFGs with fixed syntax, where parsing is unnecessary. Here, we (i) introduce a tunable class of PCFGs in which both the degree of ambiguity and the correlation structure across scales can be controlled; (ii) provide a learning mechanism -- an inference algorithm inspired by the structure of deep convolutional networks -- that links learnability and sample complexity to specific language statistics; and (iii) validate our predictions empirically across deep convolutional and transformer-based architectures. Overall, we propose a unifying framework where correlations at different scales lift local ambiguities, enabling the emergence of hierarchical representations of the data.

stat.ML↗

arXiv2025

Coarse grained descriptions of the dynamics of yielding of amorphous solids under cyclic shear

Recent computer simulations reveal several intriguing features in the evolution of properties of amorphous solids subjected to repeated cyclic shear deformation. These include the divergence of the number of cycles to reach steady states as the yielding point is approached, a non-monotonic change of properties with cycles, and the possibility of a spectrum of frozen states. Theoretical attempts to capture these properties through simple models, including the Ehrenfest model describing a random walk in a confining potential, have met partial success. Here, we show that incorporating the influence of mechanical noise through a feedback term leads to a genuine dynamical transition with characteristics reflecting those of yielding. Coarse graining the dynamics into a small number of variables leads to new insights regarding the dynamics of yielding.

cond-mat.stat-mech↗

arXiv2024

Ductile and brittle yielding of athermal amorphous solids: a mean-field paradigm beyond the random field Ising model

Developing a unified theory describing both ductile and brittle yielding constitutes a fundamental challenge of non-equilibrium statistical physics. Recently, it has been proposed that the nature of the yielding transition is controlled by physics akin to that of the quasistatically driven Random field Ising model (RFIM), which has served as the paradigm for understanding the effect of quenched disorder in slowly driven systems with short-ranged interactions. However, this theoretical picture neglects both the dynamics of, and the elasticity-induced long-ranged interactions between, the mesoscopic material constituents. Here, we address these two aspects and provide a unified theory building on the Hébraud-Lequeux elastoplastic description. The first aspect is crucial to understanding the competition between the imposed deformation rate and the finite timescale of plastic rearrangements: we provide a dynamical description of the macroscopic stress drop and predictions for the divergence of the peak susceptibility with inverse shear rate. The second is essential in order to capture properly the behaviour in the limit of quasistatic driving, where avalanches of plasticity diverge with system size at any value of the strain. We fully characterise the avalanche behaviour, which is radically different to that of the RFIM. In the quasistatic, infinite size limit, we find that both models have mean-field Landau exponents, obscuring the effect of the interactions. We show, however, that the latter profoundly affect the behaviour of finite systems approaching the spinodal-like brittle yield point, where we recover qualitatively the finite-size trends found in particle simulations, and modify the nature of the random critical point separating ductile and brittle yielding, where we predict critical behaviour on top of the marginality present at any value of the strain.

cond-mat.soft↗

arXiv2022

Mean field description of aging linear response in athermal amorphous solids

We study the linear response to strain in a mean field elastoplastic model for athermal amorphous solids, incorporating the power-law mechanical noise spectrum arising from plastic events. In the "jammed" regime of the model, where the plastic activity exhibits a non-trivial slow relaxation referred to as aging, we find that the stress relaxes incompletely to an age-dependent plateau, on a timescale which grows with material age. We determine the scaling behaviour of this aging linear response analytically, finding that key scaling exponents are universal and independent of the noise exponent $μ$. For $μ>1$, we find simple aging, where the stress relaxation timescale scales linearly with the age $t_{\rm w}$ of the material. At $μ=1$, which corresponds to interactions mediated by the physical elastic propagator, we find instead a $t_{\rm w}^{1/2}$ scaling arising from the stretched exponential decay of the plastic activity. We compare these predictions with measurements of the linear response in computer simulations of a model jammed system of repulsive soft athermal particles, during its slow dissipative relaxation towards mechanical equilibrium, and find good agreement with the theory.

cond-mat.soft↗

arXiv2022

Mean field theory of yielding under oscillatory shear

We study a mean field elastoplastic model, embedded within a disordered landscape of local yield barriers, to shed light on the behaviour of athermal amorphous solids subject to oscillatory shear. We show that the model presents a genuine dynamical transition between an elastic and a yielded state, and qualitatively reproduces the dependence on the initial degree of annealing found in particle simulations. For initial conditions prepared below the analytically derived threshold energy, we observe a non-trivial, non-monotonic approach to the yielded state. The timescale diverges as one approaches the yielding point from above, which we identify with the fatigue limit. We finally discuss the connections to brittle yielding under uniform shear.

cond-mat.soft↗

arXiv2020

Aging in a mean field elastoplastic model of amorphous solids

We construct a mean-field elastoplastic description of the dynamics of amorphous solids under arbitrary time-dependent perturbations, building on the work of Lin and Wyart [Phys. Rev. X 6, 011005 (2016)] for steady shear. Local stresses are driven by power-law distributed mechanical noise from yield events throughout the material, in contrast to the well-studied Hébraud-Lequeux model where the noise is Gaussian. We first use a mapping to a mean first passage time problem to study the phase diagram in the absence of shear, which shows a transition between an arrested and a fluid state. We then introduce a boundary layer scaling technique for low yield rate regimes, which we first apply to study the scaling of the steady state yield rate on approaching the arrest transition. These scalings are further developed to study the aging behaviour in the glassy regime, for different values of the exponent $μ$ characterizing the mechanical noise spectrum. We find that the yield rate decays as a power-law for $1<μ<2$, a stretched exponential for $μ=1$ and an exponential for $μ<1$, reflecting the relative importance of far-field and near-field events as the range of the stress propagator is varied. Comparison of the mean-field predictions with aging simulations of a lattice elastoplastic model shows excellent quantitative agreement, up to a simple rescaling of time.

cond-mat.soft↗

arXiv2018

Effects of heterogeneous social interactions on flocking dynamics

Social relationships characterize the interactions that occur within social species and may have an important impact on collective animal motion. Here, we consider a variation of the standard Vicsek model for collective motion in which interactions are mediated by an empirically motivated scale-free topology that represents a heterogeneous pattern of social contacts. We observe that the degree of order of the model is strongly affected by network heterogeneity: more heterogeneous networks show a more resilient ordered state; while less heterogeneity leads to a more fragile ordered state that can be destroyed by sufficient external noise. Our results challenge the previously accepted equivalence between the {\em static} Vicsek model and the equilibrium XY model on the network of connections, and point towards a possible equivalence with models exhibiting a different symmetry.

physics.soc-ph↗