Niels Lörch

Julian Arnold, Niels Lörch, Flemming Holtorf et al.

Despite the widespread use and success of machine-learning techniques for detecting phase transitions from data, their working principle and fundamental limits remain elusive. Here, we explain the inner workings and identify potential failure modes of these techniques by rooting popular machine-learning indicators of phase transitions in information-theoretic concepts. Using tools from information geometry, we prove that several machine-learning indicators of phase transitions approximate the square root of the system's (quantum) Fisher information from below -- a quantity that is known to indicate phase transitions but is often difficult to compute from data. We numerically demonstrate the quality of these bounds for phase transitions in classical and quantum systems.

Julian Arnold, Frank Schäfer, Niels Lörch

Machine learning has been successfully used to study phase transitions. One of the most popular approaches to identifying critical points from data without prior knowledge of the underlying phases is the learning-by-confusion scheme. As input, it requires system samples drawn from a grid of the parameter whose change is associated with potential phase transitions. Up to now, the scheme required training a distinct binary classifier for each possible splitting of the grid into two sides, resulting in a computational cost that scales linearly with the number of grid points. In this work, we propose and showcase an alternative implementation that only requires the training of a single multi-class classifier. Ideally, such multi-task learning eliminates the scaling with respect to the number of grid points. In applications to the Ising model and an image dataset generated with Stable Diffusion, we find significant speedups that closely correspond to the ideal case, with only minor deviations.

Julian Arnold, Niels Lörch

Fine-tuning LLMs on narrowly harmful datasets can lead to behavior that is broadly misaligned with respect to human values. To understand when and how this emergent misalignment occurs, we develop a comprehensive framework for detecting and characterizing rapid transitions during fine-tuning using both distributional change detection methods as well as order parameters that are formulated in plain English and evaluated by an LLM judge. Using an objective statistical dissimilarity measure, we quantify how the phase transition that occurs during fine-tuning affects multiple aspects of the model. In particular, we assess what percentage of the total distributional change in model outputs is captured by different aspects, such as alignment or verbosity, providing a decomposition of the overall transition. We also find that the actual behavioral transition occurs later in training than indicated by the peak in the gradient norm alone. Our framework enables the automated discovery and quantification of language-based order parameters, which we demonstrate on examples ranging from knowledge questions to politics and ethics.

Csaba Zsolnai, Niels Lörch, Julian Arnold

Detecting when public discourse shifts in response to major events is crucial for understanding societal dynamics. Real-world data is high-dimensional, sparse, and noisy, making changepoint detection in this domain a challenging endeavor. In this paper, we leverage neural networks for changepoint detection in news data, introducing a method based on the so-called learning-by-confusion scheme, which was originally developed for detecting phase transitions in physical systems. We train classifiers to distinguish between articles from different time periods. The resulting classification accuracy is used to estimate the total variation distance between underlying content distributions, where significant distances highlight changepoints. We demonstrate the effectiveness of this method on both synthetic datasets and real-world data from The Guardian newspaper, successfully identifying major historical events including 9/11, the COVID-19 pandemic, and presidential elections. Our approach requires minimal domain knowledge, can autonomously discover significant shifts in public discourse, and yields a quantitative measure of change in content, making it valuable for journalism, policy analysis, and crisis monitoring.