Affine Invariant Langevin Dynamics for rare-event sampling

Deepyaman Chakraborty, Ruben Harris, Rupert Klein, Guillermo Olicón-Méndez, Sebastian Reich, Claudia Schillings · 2026

Read Affine Invariant Langevin Dynamics for rare-event sampling by Deepyaman Chakraborty, Ruben Harris, Rupert Klein, Guillermo Olicón-Méndez, Sebastian Reich, Claudia Schillings (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-00107.txt.

Introduction

Deepyaman Chakraborty leads five co-authors in introducing a sampling method for estimating rare events in nonlinear systems, from engineering failures to blocked weather patterns. Their 2026 framework recasts rare events as Bayesian inverse problems built around a limit-state function whose zero set marks the event. Because that function is not smooth, they replace it with a smooth version that keeps the failure set intact.

The resulting distribution is explored by an interacting particle system whose affine invariance lets it adapt to stretched, uneven shapes, with no gradients required. Tests climb from two small examples to a point-vortex model of atmospheric blocking. In each case the method concentrates near the critical sets and yields good proposals for importance sampling. The authors present it as a robust, gradient-free option for rare-event estimation in unstable systems where plain simulation almost never visits the event of interest.

Catalog record

CREATORDeepyaman Chakraborty, Ruben Harris, Rupert Klein, Guillermo Olicón-Méndez, Sebastian Reich, Claudia Schillings
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE69 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-00107

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Affine Invariant Langevin Dynamics for rare-event sampling Deepyaman Chakraborty, Ruben Harris, Rupert Klein, Guillermo Olicón-Méndez, Sebastian Reich, Claudia Schillings Abstract We introduce an affine invariant Langevin dynamics (ALDI) framework for the efficient estimation of rare events in nonlinear dynamical systems. Rare events are formulated as Bayesian inverse problems through a nonsmooth limit-state function whose zero level set characterises the event of interest. To overcome the nondifferentiability of this function, we propose a smooth approximation that preserves the failure set and yields a posterior distribution satisfying the small-noise limit. The resulting potential is sampled by ALDI, a (derivative-free) interacting particle system whose affine invariance allows it to adapt to the local anisotropy of the posterior. We demonstrate the performance of the method across a hierarchy of benchmarks, namely two low-dimensional examples (an algebraic problem with convex geometry and a dynamical problem of saddle-type instability) and a point-vortex model for atmospheric blockings. In all cases, ALDI concentrates near the relevant near-critical sets and provides accurate proposal distributions for self-normalised importance sampling. The framework is computationally robust, potentially gradient-free, and well-suited for complex forward models with strong geometric anisotropy. These results highlight ALDI as a promising tool for rare-event estimation in unstable regimes of dynamical systems. Keywords: Bayesian inverse problems, rare-event estimation, Affine invariant Langevin dynamics, point-vortex dynamics 2020 Mathematics subject classification: 65C05, 65C30, 65C35, 62F15, 76B47 1 Introduction Rare events play a crucial role in many complex dynamical systems, ranging from transitions between metastable states in molecular dynamics [ 18 ] to critical shifts in atmospheric and oceanic circulation regimes. In geophysical contexts, such events include persistent atmospheric blockings —large-scale quasi-stationary flow patterns that interrupt the usual westerly circulation and can lead to prolonged periods of heat or precipitation extremes. Although these phenomena occur infrequently, their consequences can be severe, and quantifying their probability and underlying mechanisms remains a major scientific challenge [ 7 , 16 , 21 , 23 , 26 ] The mathematical and computational treatment of rare events requires specialised tools, since direct numerical simulation becomes infeasible due to the exponentially small probability of observing such anomalies. Classical approaches include importance sampling and splitting methods [ 2 , 5 ] , variational methods [ 15 ] , large deviation theory [ 30 ] and particle methods [ 8 ] . More recently, a probabilistic reformulation of rare-event estimation as a Bayesian inverse problem has been proposed [ 33 ] , enabling the use of sampling and inference techniques from uncertainty quantification. In this perspective, the rare event is identified with the conditioning of a prior measure on the event \{G(u)\leq 0\} , where G is a suitably chosen observable. The resulting posterior is highly concentrated in a low-probability region, making it a challenging target for standard sampling algorithms. Ensemble-based methods such as the Ensemble Kalman Filter (EnKF) have been widely used in data assimilation and inverse problems due to their computational efficiency and ability to handle high-dimensional state spaces [ 10 , 28 ] . They propagate and update an ensemble of particles according to system dynamics and observational constraints, providing a derivative-free approximation of Bayesian updates. Ensemble-based methods such as the EnKF have proved effective far beyond the linear–Gaussian setting and are routinely applied to nonlinear, non-Gaussian problems [ 9 , 10 , 28 , 31 ] . Their analysis step, however, is built around updating the first and second moments via a linear correction, which—together with finite-ensemble effects, localisation, and inflation—can bias tail probabilities and collapse separated modes unless special care is taken (e.g., tempering, iterative updates, or mixture/particle–ensemble hybrids) [ 31 , 4 ] . Consequently, representing an extremely small posterior mass in the tail region relevant for rare events remains challenging in practice, even though many EnKF variants alleviate parts of this issue. To target rare events more directly, [ 33 ] proposed an EnKF-based formulation in which the event is encoded through an observable G (taken, without loss of generality, to be nonpositive on the event), and the inverse problem \tilde{G}(u)=0,\qquad\text{with}\quad\tilde{G}(u):=\max\{0,G(u)\}, (1.1) so that the event corresponds to the set \{G(u)\leq 0\} . This casts rare-event estimation as Bayesian conditioning on (a smoothed version of) the indicator of the rare set, thereby providing an inference route that we will retain here while replacing the ensemble update with an affine-invariant Langevin sampler (ALDI) [ 14 ] as a stochastic, geometry-adapted sampling method for rare-event estimation. ALDI combines affine invariance, ensuring that the sampling dynamics is unaffected by linear rescaling or anisotropy of the state space, with the stochastic exploration of Langevin-type diffusions. This results in improved stability and efficiency, especially in high-dimensional or strongly anisotropic settings. The present work thus bridges ensemble-based inference, stochastic dynamics, and rare-event analysis within a unified geometric framework, building upon earlier developments in Bayesian sampling and dynamical systems theory. In [ 3 ] , a related method based on a combination of Langevin samplers and importance sampling (IS) has been introduced, which employs a smooth approximation of the failure event and provides an analysis of both the resulting importance sampling error and the discretization bias of the Langevin algorithm used for sampling the smoothed density. This forms the basis for an optimal computational work distribution between the Langevin sampler and the IS step. In contrast, our focus is on the small-noise limit of the associated Bayesian formulation: we characterise the limiting posterior as the prior restricted to the failure set and prove total variation convergence. This offers a measure-theoretic foundation that is independent of any specific sampling scheme and complements the algorithmic analysis in [ 3 ] . We apply ALDI to rare-event estimation in two simple test cases to analyse its behaviour, and then a point-vortex model for two-dimensional flow, used to explore the persistence of quasistationary vortex configurations associated with atmospheric blockings [ 19 ] . In both cases, we identify two qualitatively different mechanisms that generate low-probability events: • Parametric rare events , where the event \{G_{\lambda}(u)\leq 0\} occurs only for specific parameter values \lambda , and • Dynamic rare events , where the event arises from the intrinsic instability of the system’s dynamics near invariant structures. Understanding and quantifying such rare transitions is essential in many applications, as they often correspond to critical transitions or bifurcations in stochastic systems. Rare events are also related to bifurcations in stochastic dynamics when the noisy perturbations are bounded [ 20 , 29 ] . Atmospheric blockings, for instance, represent prolonged excursions of the flow dynamics toward atypical attractors, the occurrence of which is often controlled by a delicate interplay between stochastic forcing, energy input, and multiscale interactions [ 32 , 13 ] . Contributions and structure The main contributions of this paper are as follows: • Based on the reformulation of rare-event estimation as a Bayesian inverse problem, we introduce an ALDI method for exploring the associated conditional measure. • We prove that, as the observation noise tends to zero, the estimator obtained from the affine-invariant Langevin dynamics converges to the true probability of the rare event, independently of the value of the smoothing parameter. • We demonstrate the performance of ALDI on dynamical systems around unstable invariant objects, including vortex configurations related to atmospheric blockings. The remainder of

Read the complete text