Highlights

How much information about a quantum system is needed in order to control it? Maxwell's demon famously reduces the entropy of a gas by tracking its particles, but in quantum many-body systems a complete record of the state is out of reach. Giovanni Cemin and Marín Bukov of the Max-Planck Institute for the Physics of Complex Systems, together with Markus Schmitt, of the University of Regensburg, have now shown that reinforcement learning with only partial state information can stabilize nonequilibrium phases of matter, revealing information as a physical resource for controlling quantum many-body systems. The learned strategies display physically meaningful patterns and offer insights into how information can be exploited. The framework directly controls the individual dynamics of hundreds of interacting qubits, thus advancing the state-of-the-art in reinforcement learning. It establishes key connections between quantum many-body physics, control theory, and machine learning, and opens a path toward active, information-driven control of nonequilibrium quantum matter.

G. Cemin, M. Schmitt, and M. Bukov, Mach. Learn.: Sci. Technol. 7, in press (2026)

The motion of tiny self-propelled objects, from swimming microorganisms to artificial microscopic particles, can change drastically when they encounter walls. Yanis Baouche, Mathis Guéneau, and Christina Kurzthaler of the Max Planck Institute for the Physics of Complex Systems have now developed an analytical framework, based on modern techniques of statistical physics, that predicts both how long such active particles take to reach a wall and where they are likely to be found as time passes. They also show how increasing self-propulsion causes particles to spend more time near the boundaries, producing the wall accumulation seen in experiments and computer simulations. Their results provide a way to understand and predict transport in confined microscopic systems, with relevance to biological swimmers and active matter more broadly.

Y. Baouche*, M. Guéneau*, and C. Kurzthaler, Phys. Rev. Lett. 137, 138301 (2026)
* contributed equally

The application deadline for this call is the 13 of November 2026. Distinguished PKS postdoctoral fellows appear personally along with the departments and groups on the main research page of the institute and are expected to have at least one year of postdoctoral experience at an institution other than the one at which their PhD was awarded. Applications for this fellowship directly after completion of the PhD might be considered in exceptional cases. Please click on the button to see the full advertisement!

How can a sensing system balance the information it gains against the energy it spends when the underlying processes are only partially observable? Giorgio Nicoletti, Ivan di Terlizzi, and Daniel M. Busiello of the Max Planck Institute for the Physics of Complex Systems, the Abdus Salam International Center for Theoretical Physics (Trieste), the Ludwig-Maximilians-Universität Munich, and the University of Padova have introduced a framework in which observable fluctuations can be used to estimate information and dissipation and adapt sensing strategies on the fly. The work connects stochastic thermodynamics with biological adaptation and provides new tools for understanding information processing in biochemical and synthetic sensing systems.

G. Nicoletti et al., Phys. Rev. Lett. 137, 127101 (2026), highlighted as an Editors' Suggestion

Selected for a Viewpoint: Physics 19, 121 (2026)

Marko Popović has received the Pierre-Gilles de Gennes Prize, awarded as a part of the From Solid State to Biophysics conference series in Cavtat, Croatia. The prize recognises mid-career researchers who bridge the gap between solid-state physics or nanoscience and the fields of biophysics and life sciences. The award commemorates Nobel Laureate Pierre-Gilles de Gennes, whose work connected ideas from solid-state physics, soft matter, and biological systems. Marko received the prize in recognition of his research at the interface of physics and biology, including his work on the mechanics and organization of living tissues.

The discovery and understanding of new forms of superconducting pairing states remain central challenges in quantum science, as many exotic pairing states are difficult to realise, characterise, and connect to experimentally measurable signals. In particular, superconducting phases with non-zero momentum pairing are theoretically intriguing but challenging to identify in realistic strongly interacting systems.
Chunhan Feng of the Max Planck Institute for the Physics of Complex Systems, together with collaborators at the Flatiron Institute and MIT, has now developed approaches to identify exotic pairing states in strongly interacting quantum systems. The researchers investigated a special type of superconducting state in which pairs of particles form with a finite momentum, leading to a repeating spatial pattern that goes beyond the conventional picture of superconductivity.
By combining large-scale many-body simulations with experimentally accessible measurements in ultracold-atom systems, they mapped out the parameter regimes in which this exotic pairing may emerge and showed that its early signatures could be detected with current experimental techniques. This work bridging the gap between theory and experiment thus provides a pathway to uncover previously hidden quantum phases of matter.

C. Feng et al., Proc. Natl. Acad. Sci. USA 123, e2600303123 (2026)

Image generated using ChatGPT.

Birds in flocks, bacteria, cells: in many collective systems, individual elements respond to only part of their surroundings, seemingly defying Newton’s third law of motion, that action equals reaction. These exceptions are known as non-reciprocal interactions. Even though these interactions do not stem from a potential, researchers at the Max Planck Institute for the Physics of Complex Systems have now developed a way to capture them in a Hamiltonian. The trick that made this possible is to introduce a fictitious partner for each component of the system to replace non-reciprocal interactions with reciprocal interactions with auxiliary degrees of freedom. This work advances the study of non-reciprocal systems significantly, because it enables use the established theoretical frameworks of many-body physics to describe these systems, which is an important step towards building the statistical mechanics of non-reciprocal systems.

Yu-Bo Shi, Roderich Moessner, Ricard Alert, and Marín Bukov, Nat. Phys. (2026)

Image credit: Kilian Neddermeyer

This highlight is based, in part, on the text of a press release by the ctd.qmat Cluster of Excellence (Universität Würzburg and TU Dresden), which is available here.

Controlling how fast quantum systems relax toward their steady state is a central challenge for quantum simulation and state preparation, especially in the presence of dissipation. In a joint effort between the Biological Physics and Condensed Matter divisions of the Max Planck Institute for the Physics of Complex Systems, Nicolò Beato and Gianluca Teza have now introduced a general strategy to tune relaxation dynamics in open quantum systems.
Their approach is inspired by ideas related to the Mpemba effect—the counterintuitive observation that hot water can sometimes freeze faster than cold. Extending this concept to the quantum realm, the authors show that relaxation can be dramatically altered by preparing the system in a carefully chosen initial state. By applying a single unitary transformation, the method selectively suppresses specific dissipative modes that would otherwise slow down (or speed up) the dynamics. Unlike conventional control protocols, this strategy requires no active intervention during the evolution itself and remains effective even in systems with dense and complex relaxation spectra.
By enabling both acceleration and prolongation of relaxation under realistic constraints, this work opens new avenues for controlling nonequilibrium quantum dynamics in modern quantum simulators.

Figure: A unitary transformation acts as a spectral filter for dissipation in an open quantum system. By selectively redirecting slow relaxation modes while preserving the steady state, the transformation reshapes the relaxation dynamics without requiring active control during evolution, enabling precise tuning even in the presence of dense dissipative spectra.

Nicoló Beato and Gianluca Teza, Phys. Rev. Lett. 136, 070401 (2026)

Nearly three centuries ago, Euler explained how an elastic line buckles under compression in the plane. This classical, minimal calculation epitomises mechanical bifurcations.
Shiheng Zhao and Pierre Haas of the Max Planck Institute for the Physics of Complex Systems and the Max Planck Institute of Molecular Cell Biology and Genetics have now answered the question that Euler did not solve: How does such a compressed line buckle within a curved surface? Combining exact and numerical calculations, the researchers discovered that the mathematical bifurcation structure changes fundamentally. In particular, there can be a surprising "snap-through" instability at larger compressions, whereby the elastic line suddenly, discontinuously jumps to a higher compression. These theoretical results are the foundations explaining a class of elastic instabilities within curved surfaces that also has biological relevance: for example, in a collaboration with researchers at Princeton and the Flatiron Institute, Shiheng Zhao and Pierre Haas recently showed that a similar mechanical bifurcation within a curved surface underpins the morphogenesis of the hindgut of the fruitfly Drosophila.

Shiheng Zhao and Pierre A. Haas, Phys. Rev. Lett. 135, 247201 (2025)

Understanding complex many-body dynamics in laser-driven molecules is essential for efforts to control chemical reactions using intense light fields.
Ultrashort, high-intensity X-ray pulses from accelerator-based free-electron lasers (FELs) now make it possible to observe—in real time—how strong laser fields reshape molecular structure. Physicists from two Max Planck Institutes, the MPI for Nuclear Physics in Heidelberg and the MPI for the Physics of Complex Systems in Dresden, together with collaborators from the Max Born Institute in Berlin and institutions in Switzerland, the USA, and Japan, have investigated the prototype molecule C₆₀, the iconic football-shaped "Buckminsterfullerene". In this study, an intense femtosecond laser pulse initiates the molecular dynamics, and a subsequent femtosecond X-ray pulse images the evolving structure at a precisely controlled time delay (as illustrated in the figure).
The experiments were conducted at the Linac Coherent Light Source (LCLS) at SLAC National Accelerator Laboratory, an X-ray FEL operated by Stanford University. Only the enormous intensities provided by such facilities make these imaging experiments possible. Since multi-electron dynamics under intense laser irradiation remain extremely challenging to describe theoretically, with a full quantum-mechanical treatment still out of reach, such X-ray "movies" of structural dynamics offer an ideal testbed for advancing our understanding of fundamental quantum processes in increasingly large and complex molecular systems.

Kirsten Schnorr, Sven Augustin, Ulf Saalmann et al., Sci. Adv. 11, eadz1900 (2025)