In-context learning to predict critical transitions in dynamical systems

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Sevinchan, Yunus, Nathaniel, Juan, Ueltzhöffer, Kai, Roesch, Carla, Weber, Tobias, Laschos, Vaios, Fan, Hang, Ramien, Gregor, Haux, Johannes, Gentine, Pierre, Herdeanu, Benjamin
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910213114691584
author Sevinchan, Yunus
Nathaniel, Juan
Ueltzhöffer, Kai
Roesch, Carla
Weber, Tobias
Laschos, Vaios
Fan, Hang
Ramien, Gregor
Haux, Johannes
Gentine, Pierre
Herdeanu, Benjamin
author_facet Sevinchan, Yunus
Nathaniel, Juan
Ueltzhöffer, Kai
Roesch, Carla
Weber, Tobias
Laschos, Vaios
Fan, Hang
Ramien, Gregor
Haux, Johannes
Gentine, Pierre
Herdeanu, Benjamin
contents Critical transitions - abrupt, often irreversible changes in system dynamics - arise across human and natural systems, often with catastrophic consequences. Real-world observations of such shifts remain scarce, preventing the development of reliable early warning systems. Conventional statistical and spectral indicators, such as increasing variance, tend to fail under realistic conditions of limited data and correlated noise, whereas existing deep learning classifiers do not extrapolate beyond their training data distribution. In this work, we introduce TipPFN, an in-context learning (ICL) framework that uses a prior-data fitted network to infer a system's proximity to a critical transition. Trained on our novel synthetic data generator, which is based on canonical bifurcation scenarios coupled to diverse, randomized stochastic dynamics, TipPFN flexibly capitalizes on contexts of various sizes, complexity and dimensionalities. We demonstrate robust, state-of-the-art early detection of critical transitions in previously unseen tipping regimes, sim-to-real examples, and real-world observations in both ICL and zero-shot settings.
format Preprint
id arxiv_https___arxiv_org_abs_2605_12308
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle In-context learning to predict critical transitions in dynamical systems
Sevinchan, Yunus
Nathaniel, Juan
Ueltzhöffer, Kai
Roesch, Carla
Weber, Tobias
Laschos, Vaios
Fan, Hang
Ramien, Gregor
Haux, Johannes
Gentine, Pierre
Herdeanu, Benjamin
Machine Learning
Critical transitions - abrupt, often irreversible changes in system dynamics - arise across human and natural systems, often with catastrophic consequences. Real-world observations of such shifts remain scarce, preventing the development of reliable early warning systems. Conventional statistical and spectral indicators, such as increasing variance, tend to fail under realistic conditions of limited data and correlated noise, whereas existing deep learning classifiers do not extrapolate beyond their training data distribution. In this work, we introduce TipPFN, an in-context learning (ICL) framework that uses a prior-data fitted network to infer a system's proximity to a critical transition. Trained on our novel synthetic data generator, which is based on canonical bifurcation scenarios coupled to diverse, randomized stochastic dynamics, TipPFN flexibly capitalizes on contexts of various sizes, complexity and dimensionalities. We demonstrate robust, state-of-the-art early detection of critical transitions in previously unseen tipping regimes, sim-to-real examples, and real-world observations in both ICL and zero-shot settings.
title In-context learning to predict critical transitions in dynamical systems
topic Machine Learning
url https://arxiv.org/abs/2605.12308