arXiv subjects

Léo Touzo

Publications and source records attributed to Léo Touzo.

12 recordsLinked to original sources

arXiv2026

Wave turbulence theory of odd fluids and solids: kinetic equations and solutions

The theory of wave turbulence describes the properties of physical systems composed of a set of weak-amplitude random waves interacting nonlinearly. Here, we study odd wave turbulence, which arises in chiral media subjected to non-reciprocal stresses, notably odd viscosity and odd elasticity. In both cases, we consider simple models for which we can derive and solve analytically the kinetic equations describing the long-term statistical behavior of spectral quantities such as energy or wave action. For odd viscosity, we consider a three-dimensional model that exhibits wave turbulence involving three-wave interactions, which gives rise to a direct energy cascade characterized by an anisotropic Kolmogorov-Zakharov (KZ) spectrum. For odd elasticity, we consider a quasi-one-dimensional overdamped model that exhibits much slower dynamics involving six-wave interactions. In that case, the KZ spectrum corresponding to a forward cascade of a conserved quantity we call odd energy, is nonlocal and therefore does not constitute a physical solution. However, the other KZ solution, which describes an inverse cascade of wave action, is only marginally non-local and is therefore valid up to a logarithmic correction. These two analytical theories provide a rigorous interpretation of direct numerical simulations, where the KZ spectrum is observed both in the case of odd viscosity (forward cascade) and of odd elasticity (inverse cascade).

physics.flu-dyn↗

arXiv2026

Odd wave turbulence

Wave turbulence describes the statistical dynamics of weakly interacting waves in out-of-equilibrium systems. These systems are out of equilibrium because forcing and dissipation occur at different scales, leading to a flux across scales known as a turbulent cascade. However, the underlying medium is usually assumed to be passive. Here, we study situations where a turbulent cascade takes place on top of a driven-dissipative steady-state, which is already out of equilibrium without forcing. We focus on chiral active media, in which waves arise as a consequence of non-reciprocal responses known as odd viscosity and odd elasticity. Combining analytical theory with large-scale direct numerical simulations, we show that fluids with odd viscosity sustain weak wave turbulence down to the smallest active scales, leading to anisotropic Kolmogorov-Zakharov spectra and a modified forward cascade. In the odd elastic solids, two new conserved quantities emerge (which we call wave action and odd energy), giving rise to an inverse cascade in the weak regime and a forward cascade in the strong regime. Our results pave the way towards understanding the chaotic nonlinear regime of systems with nonreciprocal responses from chiral plasma to engineered solids.

physics.flu-dyn↗

arXiv2026

Non-uniqueness of the steady state for run-and-tumble particles with a double-well interaction potential

We study $N$ run-and-tumble particles (RTPs) in one dimension interacting via a double-well potential $W(r)=-k_0 \, r^2/2+g \, r^4/4$, which is repulsive at short interparticle distance $r$ and attractive at large distance. At large time, the system forms a bound state where the density of particles has a finite support. We focus on the determination of the total density of particles in the stationary state $ρ_s(x)$, in the limit $N\to+\infty$. We obtain an explicit expression for $ρ_s(x)$ as a function of the ''renormalized" interaction parameter $k=k_0-3m_2$ where $m_2$ is the second moment of $ρ_s(x)$. Interestingly, this stationary solution exhibits a transition between a connected and a disconnected support for a certain value of $k$, which has no equivalent in the case of Brownian particles. Analyzing in detail the expression of the stationary density in the two cases, we find a variety of regimes characterized by different behaviors near the edges of the support and around $x=0$. Furthermore, we find that the mapping $k_0\to k$ becomes multi-valued below a certain value of the tumbling rate $γ$ of the RTPs for some range of values of $k_0$ near the transition, implying the existence of two stable solutions. Finally, we show that in the case of a disconnected support, it is possible to observe steady states where the density $ρ_s(x)$ is not symmetric. All our analytical predictions are in good agreement with numerical simulations already for systems of $N = 100$ particles. The non-uniqueness of the stationary state is a particular feature of this model in the presence of active (RTP) noise, which contrasts with the uniqueness of the Gibbs equilibrium for Brownian particles. We argue that these results are also relevant for a class of more realistic interactions with both an attractive and a repulsive part, but which decay at infinity.

cond-mat.stat-mech↗

arXiv2025

Exact results for active particle models: from long-range interactions to first-passage properties

The goal of this thesis is to obtain new exact results for models of active particles in one dimension, focusing on two different aspects: their behavior in the presence of long-range interactions and their first-passage properties. In the first part we give an overview of existing exact results both for active particle models and for Brownian particles with long-range interactions (Riesz gases). The next two parts focus on how methods from these two fields can be combined and extended to derive new results for models of active particles with long-range interactions. In part two, we study the density of particles in the stationary state, in the limit where the number of particles is very large, using an extension of the Dean-Kawasaki equation to run-and-tumble particles (RTPs). In the case of the 1D Coulomb interaction (attractive or repulsive), we obtain exact expressions for the stationary density for different types of confining potentials, which sheds lights on new non-equilibrium phase-transitions. Some results are also obtained for a repulsive 2D Coulomb interaction (log-gas), although the single-file constraint makes the study more difficult in this case. In part three, we focus on the fluctuations at the tagged particle level. In the limit of weak noise, we compute exactly and analyze in different regimes a variety of correlation functions of the particle positions and interparticle distances, both for the Brownian Riesz gas and for its active counterpart, and show that the activity plays an important role both at short times and at small distances. The last part of this thesis focuses on Siegmund duality, which connects the first-passage properties of a stochastic process with absorbing boundaries to its spatial distribution with hard walls. We extend this duality to a new class of stochastic processes, which includes active particles and diffusing diffusivity models.

cond-mat.stat-mech↗

arXiv2025

Run-and-tumble particles with 1D Coulomb interaction: the active jellium model and the non-reciprocal self-gravitating gas

Recently we studied $N$ run-and-tumble particles in one dimension - which switch with rate $γ$ between driving velocities $\pm v_0$ - interacting via the long range 1D Coulomb potential (also called rank interaction), both in the attractive and in the repulsive case, with and without a confining potential. We extend this study in two directions. First we consider the same system, but inside a harmonic confining potential, which we call "active jellium". We obtain a parametric representation of the particle density in the stationary state at large $N$, which we analyze in detail. Contrary to the linear potential, there is always a steady-state where the density has a bounded support. However, we find that the model still exhibits transitions between phases with different behaviors of the density at the edges, ranging from a continuous decay to a jump, or even a shock (i.e. a cluster of particles, which manifests as a delta peak in the density). Notably, the interactions forbid a divergent density at the edges, which may occur in the non-interacting case. In the second part, we consider a non-reciprocal version of the rank interaction: the $+$ particles (of velocity $+v_0$) are attracted towards the $-$ particles (of velocity $-v_0$) with a constant force $b/N$, while the $-$ particles are repelled by the $+$ particles with a force of same amplitude. In order for a stationary state to exist we add a linear confining potential. We derive an explicit expression for the stationary density at large $N$, which exhibits an explicit breaking of the mirror symmetry with respect to $x=0$. This again shows the existence of several phases, which differ by the presence or absence of a shock at $x=0$, with one phase even exhibiting a vanishing density on the whole region $x>0$. Our analytical results are complemented by numerical simulations for finite $N$.

cond-mat.stat-mech↗

arXiv2024

Siegmund duality for physicists: a bridge between spatial and first-passage properties of continuous and discrete time stochastic processes

We consider a generic one-dimensional stochastic process $x(t)$, or a random walk $X_n$, which describes the position of a particle evolving inside an interval $[a,b]$, with absorbing walls located at $a$ and $b$. In continuous time, $x(t)$ is driven by some equilibrium process $\mathbfθ(t)$, while in discrete time, the jumps of $X_n$ follow a stationary process that obeys a time reversal property. An important observable to characterize its behaviour is the exit probability $E_b(x,t)$, which is the probability for the particle to be absorbed first at the wall $b$, before or at time $t$, given its initial position $x$. In this paper we show that the derivation of this quantity can be tackled by studying a dual process $y(t)$ very similar to $x(t)$ but with hard walls at $a$ and $b$. More precisely, we show that the quantity $E_b(x,t)$ for the process $x(t)$ is equal to the probability $\tilde Φ(x,t|b)$ of finding the dual process inside the interval $[a,x]$ at time $t$, with $y(0) =b$. This is known as Siegmund duality in mathematics. Here we show that this duality applies to various processes which are of interest in physics, including models of active particles, diffusing diffusivity models, a large class of discrete and continuous time random walks, and even processes subjected to stochastic resetting. For all these cases, we provide an explicit construction of the dual process. We also give simple derivations of this identity both in the continuous and in the discrete time setting, as well as numerical tests for a large number of models of interest. Finally, we use simulations to show that the duality is also likely to hold for more complex processes such as fractional Brownian motion.

cond-mat.stat-mech↗

arXiv2024

Relating absorbing and hard wall boundary conditions for a one-dimensional run-and-tumble particle

The connection between absorbing boundary conditions and hard walls is well established in the mathematical literature for a variety of stochastic models, including for instance the Brownian motion. In this paper we explore this duality for a different type of process which is of particular interest in physics and biology, namely the run-tumble-particle, a toy model of active particle. For a one-dimensional run-and-tumble particle subjected to an arbitrary external force, we provide a duality relation between the exit probability, i.e. the probability that the particle exits an interval from a given boundary before a certain time $t$, and the cumulative distribution of its position in the presence of hard walls at the same time $t$. We show this relation for a run-and-tumble particle in the stationary state by explicitly computing both quantities. At finite time, we provide a derivation using the Fokker-Planck equation. All the results are confirmed by numerical simulations.

cond-mat.stat-mech↗

arXiv2024

Abrupt and spontaneous strategy switches emerge in simple regularised neural networks

Humans sometimes have an insight that leads to a sudden and drastic performance improvement on the task they are working on. Sudden strategy adaptations are often linked to insights, considered to be a unique aspect of human cognition tied to complex processes such as creativity or meta-cognitive reasoning. Here, we take a learning perspective and ask whether insight-like behaviour can occur in simple artificial neural networks, even when the models only learn to form input-output associations through gradual gradient descent. We compared learning dynamics in humans and regularised neural networks in a perceptual decision task that included a hidden regularity to solve the task more efficiently. Our results show that only some humans discover this regularity, whose behaviour was marked by a sudden and abrupt strategy switch that reflects an aha-moment. Notably, we find that simple neural networks with a gradual learning rule and a constant learning rate closely mimicked behavioural characteristics of human insight-like switches, exhibiting delay of insight, suddenness and selective occurrence in only some networks. Analyses of network architectures and learning dynamics revealed that insight-like behaviour crucially depended on a regularised gating mechanism and noise added to gradient updates, which allowed the networks to accumulate "silent knowledge" that is initially suppressed by regularised (attentional) gating. This suggests that insight-like behaviour can arise naturally from gradual learning in simple neural networks, where it reflects the combined influences of noise, gating and regularisation.

cs.AI↗

arXiv2023

Non-equilibrium phase transitions in active rank diffusions

We consider N run and tumble particles in one dimension interacting via a linear 1D Coulomb potential, an active version of the rank diffusion problem. It was solved previously for N = 2 leading to a stationary bound state in the attractive case. Here the evolution of the density fields is obtained in the large N limit in terms of two coupled Burger's type equations. In the attractive case the exact stationary solution describes a non-trivial N-particle bound state, which exhibits transitions between a phase where the density is smooth with infinite support, a phase where the density has finite support and exhibits "shocks", i.e. clusters of particles, at the edges, and a fully clustered phase. In presence of an additional linear potential, the phase diagram, obtained for either sign of the interaction, is even richer, with additional partially expanding phases, with or without shocks. Finally, a general self-consistent method is introduced to treat more general interactions. The predictions are tested through extensive numerical simulations.

cond-mat.stat-mech↗

arXiv2022

Branching with selection and mutation I: Mutant fitness of Fréchet type

We investigate two stochastic models of a growing population subject to selection and mutation. In our models each individual carries a fitness which determines its mean offspring number. Many of these offspring inherit their parent's fitness, but some are mutants and obtain a fitness randomly sampled from a distribution in the domain of attraction of the Fréchet distribution. We give a rigorous proof for the precise rate of superexponential growth of these stochastic processes and support the argument by a heuristic and numerical study of the mechanism underlying this growth.

math.PR↗

arXiv2021

Information thermodynamics of financial markets: the Glosten-Milgrom model

The Glosten-Milgrom model describes a single asset market, where informed traders interact with a market maker, in the presence of noise traders. We derive an analogy between this financial model and a Szilárd information engine by {\em i)} showing that the optimal work extraction protocol in the latter coincides with the pricing strategy of the market maker in the former and {\em ii)} defining a market analogue of the physical temperature from the analysis of the distribution of market orders. Then we show that the expected gain of informed traders is bounded above by the product of this market temperature with the amount of information that informed traders have, in exact analogy with the corresponding formula for the maximal expected amount of work that can be extracted from a cycle of the information engine. This suggests that recent ideas from information thermodynamics may shed light on financial markets, and lead to generalised inequalities, in the spirit of the extended second law of thermodynamics.

cond-mat.stat-mech↗

arXiv2020

Optimal Work Extraction and the Minimum Description Length Principle

We discuss work extraction from classical information engines (e.g., Szilárd) with $N$-particles, $q$ partitions, and initial arbitrary non-equilibrium states. In particular, we focus on their {\em optimal} behaviour, which includes the measurement of a set of quantities $Φ$ with a feedback protocol that extracts the maximal average amount of work. We show that the optimal non-equilibrium state to which the engine should be driven before the measurement is given by the normalised maximum-likelihood probability distribution of a statistical model that admits $Φ$ as sufficient statistics. Furthermore, we show that the minimax universal code redundancy $\mathcal{R}^*$ associated to this model, provides an upper bound to the work that the demon can extract on average from the cycle, in units of $k_{\rm B}T$. We also find that, in the limit of $N$ large, the maximum average extracted work cannot exceed $H[Φ]/2$, i.e. one half times the Shannon entropy of the measurement. Our results establish a connection between optimal work extraction in stochastic thermodynamics and optimal universal data compression, providing design principles for optimal information engines. In particular, they suggest that: (i) optimal coding is thermodynamically efficient, and (ii) it is essential to drive the system into a critical state in order to achieve optimal performance.

cond-mat.stat-mech↗