arXiv subjects

Miriam Klopotek

Publications and source records attributed to Miriam Klopotek.

9 recordsLinked to original sources

arXiv2026

Entropy production of active matter systems as indicator for computing performance

Physical systems can process information through their natural dynamics, offering alternatives to conventional digital computing. Reservoir computing offers a basic framework by using a nonlinear substrate to map inputs into rich dynamical states read out by a simple linear layer. Active matter substrates are striking examples; they continuously consume energy and produce entropy. Theoretically, entropy production (EP) can describe the irreversibility and distance from equilibrium. But it remains unclear whether it can track computational capabilities. We address this conceptual gap by analyzing a driven swarm reservoir model. The system EP is computed from phase-space contraction and the bath EP from heat flow, separately, and put in direct association to prediction performance on a Lorenz-63 task. Via force parameter scans, we show that dynamical regimes with the strongest response to a driver as well as dissipation coincide with peak performance. Therein, the dynamical discrepancy between innate (minimal dissipation) and driven transferred heat (maximal dissipation) is sharpest. Generally, driver work and relative differences of driven and undriven EP closely mirror the performance landscape. The system EP, derived from a generalized Liouville-equation estimator, and heat flow provide complementary diagnostics and metrics, which are most robust in the best-performing regime. These results extend prior expectations that dissipation matters for computation by identifying when and how it becomes predictive. They also relate inference power to innate dynamics, pointing to generic principles for physical computing and where EP offers a screening metric for reservoirs and other base substrates.

cond-mat.stat-mech↗

arXiv2026

Interpreting learning dynamics of autoencoders: Transient scaling and emerging concepts of the Ising model

We study how unsupervised autoencoders trained on microscopic spin configurations from the Ising model learn macroscopic, theory-relevant variables underlying the data-generating process. We quantify learning across multiple spatial (coarse-graining) scales and reveal two distinct dynamical regimes that appear sequentially, controlled by the main hyperparameters (model depth, width, and learning rate): one in which magnetization and another in which energy is learned across scales. The first exhibits error fluctuations ordered to scale and learns global averages only; The second gradually resolves smaller scales relevant for the energy representation. Deep models trained at moderate and fast rates become arrested before reaching these regimes. We connect reconstruction errors with the latent representations using a novel analysis of self-recursive trajectories. These intrinsic dynamics are induced by prediction errors, exposing how training drives representation changes for macroscopic concepts. We utilize the intuition that learning operates as a process driven far from equilibrium by fluctuations from the training data to provide an interpretive basis grounded in both the physical world and the machine models that represent it.

cs.LG↗

arXiv2026

Capturing reduced-order quantum many-body dynamics out of equilibrium via neural ordinary differential equations

Out-of-equilibrium quantum many-body systems exhibit rapid correlation buildup that underlies many emerging phenomena. Exact wave-function methods to describe this scale exponentially with particle number; simpler mean-field approaches neglect essential two-particle correlations. The time-dependent two-particle reduced density matrix (TD2RDM) formalism offers a middle ground by propagating the two-particle reduced density matrix (2RDM) and closing the BBGKY hierarchy with a reconstruction of the three-particle cumulant. But the validity and existence of time-local reconstruction functionals ignoring memory effects remain unclear across different dynamical regimes. We show that a neural ODE model trained on exact 2RDM data (no dimensionality reduction) can reproduce its dynamics without any explicit three-particle information -- but only in parameter regions where the Pearson correlation between the two- and three-particle cumulants is large. In the anti-correlated or uncorrelated regime, the neural ODE fails, indicating that no simple time-local functional of the instantaneous two-particle cumulant can capture the evolution. The magnitude of the time-averaged three-particle-correlation buildup appears to be the primary predictor of success: For a moderate correlation buildup, both neural ODE predictions and existing TD2RDM reconstructions are accurate, whereas stronger values lead to systematic breakdowns. These findings pinpoint the need for memory-dependent kernels in the three-particle cumulant reconstruction for the latter regime. Our results place the neural ODE as a model-agnostic diagnostic tool that maps the regime of applicability of cumulant expansion methods and guides the development of non-local closure schemes. More broadly, the ability to learn high-dimensional RDM dynamics from limited data opens a pathway to fast, data-driven simulation of correlated quantum matter.

cs.LG↗

arXiv2026

Optimal information injection and transfer mechanisms for active matter reservoir computing

Reservoir computing (RC) is a state-of-the-art machine learning method that makes use of the power of dynamical systems (the reservoir) for real-time inference. When using biological complex systems as reservoir substrates, it serves as a testbed for basic questions about bio-inspired computation -- of how self-organization generates proper spatiotemporal patterning. Here, we use a simulation of an active matter system, driven by a chaotically moving input signal, as a reservoir. So far, it has been unclear whether such complex systems possess the capacity to process information efficiently and independently of the method by which it was introduced. We find that when switching from a repulsive to an attractive driving force, the system completely changes the way it computes, while the predictive performance landscapes remain nearly identical. The nonlinearity of the driver's injection force improves computation by decoupling the single-agent dynamics from that of the driver. Triggered are the (re-)growth, deformation, and active motion of smooth structural boundaries (interfaces), and the emergence of coherent gradients in speed -- features found in many soft materials and biological systems. The nonlinear driving force activates emergent regulatory mechanisms, which manifest enhanced morphological and dynamic diversity -- arguably improving fading memory, nonlinearity, expressivity, and thus, performance. We further perform RC in a broad variety of non-equilibrium active matter phases that arise when tuning internal (repulsive) forces for information transfer. Overall, we find that active matter agents forming liquid droplets are particularly well suited for RC. The consistently convex shape of the predictive performance landscapes, together with the observed phenomenological richness, conveys robustness and adaptivity.

nlin.AO↗

arXiv2026

Robustly optimal dynamics for active matter reservoir computing

Information processing abilities of active matter are studied in the reservoir computing (RC) paradigm to infer the future state of a chaotic signal. We uncover an exceptional regime of agent dynamics that has been overlooked previously. It appears robustly optimal for performance under many conditions, thus providing valuable insights into computation with physical systems more generally. The key to forming effective mechanisms for information processing appears in the system's intrinsic relaxation abilities. These are probed without actually enforcing a specific inference goal. The dynamical regime that achieves optimal computation is located just below a critical damping threshold, involving a relaxation with multiple stages, and is readable at the single-particle level. At the many-body level, it yields substrates robustly optimal for RC across varying physical parameters and inference tasks. A system in this regime exhibits a strong diversity of dynamic mechanisms under highly fluctuating driving forces. Correlations of agent dynamics can express a tight relationship between the responding system and the fluctuating forces driving it. As this model is interpretable in physical terms, it facilitates re-framing inquiries regarding learning and unconventional computing with a fresh rationale for many-body physics out of equilibrium.

nlin.AO↗

arXiv2025

Interpretable Machine Learning in Physics: A Review

Machine learning is increasingly transforming various scientific fields, enabled by advancements in computational power and access to large data sets from experiments and simulations. As artificial intelligence (AI) continues to grow in capability, these algorithms will enable many scientific discoveries beyond human capabilities. Since the primary goal of science is to understand the world around us, fully leveraging machine learning in scientific discovery requires models that are interpretable -- allowing experts to comprehend the concepts underlying machine-learned predictions. Successful interpretations increase trust in black-box methods, help reduce errors, allow for the improvement of the underlying models, enhance human-AI collaboration, and ultimately enable fully automated scientific discoveries that remain understandable to human scientists. This review examines the role of interpretability in machine learning applied to physics. We categorize different aspects of interpretability, discuss machine learning models in terms of both interpretability and performance, and explore the philosophical implications of interpretability in scientific inquiry. Additionally, we highlight recent advances in interpretable machine learning across many subfields of physics. By bridging boundaries between disciplines -- each with its own unique insights and challenges -- we aim to establish interpretable machine learning as a core research focus in science.

physics.comp-ph↗

arXiv2024

Machine learning stochastic differential equations for the evolution of order parameters of classical many-body systems in and out of equilibrium

We develop a machine learning algorithm to infer the emergent stochastic equation governing the evolution of an order parameter of a many-body system. We train our neural network to independently learn the directed force acting on the order parameter as well as an effective diffusive noise. We illustrate our approach using the classical Ising model endowed with Glauber dynamics, and the contact process as test cases. For both models, which represent paradigmatic equilibrium and nonequilibrium scenarios, the directed force and noise can be efficiently inferred. The directed force term of the Ising model allows us to reconstruct an effective potential for the order parameter which develops the characteristic double-well shape below the critical temperature. Despite its genuine nonequilibrium nature, such an effective potential can also be obtained for the contact process and its shape signals a phase transition into an absorbing state. Also, in contrast to the equilibrium Ising model, the presence of an absorbing state renders the noise term dependent on the value of the order parameter itself.

cond-mat.dis-nn↗

arXiv2021

Lattice gas study of thin film growth scenarios and transitions between them: Role of substrate

Thin film growth is investigated in two types of lattice gas models where substrate and film particles are different, expressed by unequal interaction energy parameters. The first is of solid-on-solid type, whereas the second type incorporates desorption, diffusion in the gas phase above the film and re-adsorption at the film (appropriate for growth in colloidal systems). The difference between particle-substrate and particle-particle interactions plays a central role for the evolution of the film morphology at intermediate times. The models exhibit a dynamic layering transition which occurs at generally lower substrate attraction strengths than the equilibrium layering transition. A second, flattening transition is found where initial island growth transforms to layer-by-layer growth at intermediate deposition times. Combined with the known roughening behavior in such models for very large deposition times, we present four global growth scenarios, charting out the possible types of roughness evolution.

cond-mat.soft↗

arXiv2020

Strategies for the reuse of N95 and N99 respiratory masks during the COVID-19 pandemic

The COVID-19 pandemic presents a strain of unprecedented scale on health systems around the world. In order to reliably protect medical personnel, and thus to contain the spread of the pandemic, it is essential to provide N95 or N99 (European FFP2 or FFP3) respiratory masks (FFRs). Such masks are currently in extreme shortage: To guarantee their supply sufficiently and for all cases, it would absolutely necessary to reuse them. In recent years, the scientific literature has laid out various possibilities to disinfect the FFRs for their reuse several times. We identify the most promising disinfection methods for the current critical situation (internationally) and project further methods and modifications beyond these.

physics.med-ph↗