arXiv subjects

Gu Zhang

Publications and source records attributed to Gu Zhang.

At least 19 recordsLinked to original sources

arXiv2026

Can out-of-equilibrium linear response reveal anyonic statistics?

Determination of the anyonic statistical braiding phase has relied on either Aharonov-Bohm inter- ference experiments or cross-correlations measured in collider platforms. Here we consider collider setups, supporting stationary far-from-equilibrium anyonic beams, characterized by an effective chemical potential and an effective temperature, which carry universal information about the ele- mentary charge and the statistical braiding phase. We develop an out-of-equilibrium linear response theory designed to describe charge and thermal quantum transport under deviations from this ef- fective equilibrium (unlike linear response designed to describe transport near genuine equilibrium). Our effective linear-response transport coefficients directly reflect the fractional charge and statistics of the anyons involved, avoiding the need to measure higher-order current correlations. Moreover, the emergence of finite thermoelectric (Peltier and Seebeck) coefficients signifies the presence of real anyon collisions (as opposed to virtual braiding in the time domain), intimately associated with the breaking of a characteristic particle-hole symmetry specific to anyonic gases.

cond-mat.mes-hall↗

arXiv2026

Exact fluctuation relations in voltage- and temperature-biased Laughlin-edge constrictions

We present a comprehensive analysis of non-equilibrium fluctuation-dissipation relations connect- ing experimentally accessible chiral-current auto- and cross-correlations to the tunneling-current noise and conductance in Laughlin edge states coupled through a quantum point contact (QPC). We examine their validity for two chiral Laughlin edges held at different temperatures and voltages and show that the relations remain exact for arbitrary tunneling strength, voltage bias, and edge- state temperatures. We further generalize them to spatially extended QPCs and to tunneling am- plitudes with an explicit voltage dependence, and discuss the conditions and limitations associated with these generalizations. Our results establish that the local tunneling-current noise generated at the QPC can be reliably reconstructed from experimentally accessible auto- and cross-correlations measured downstream, providing a robust route to characterize non-equilibrium transport in chiral edge states.

cond-mat.mes-hall↗

arXiv2026

SLIM-0.5B: Learning Action-Grounded Predictive Latents for Robot Manipulation

Vision-language-action policies rely on large multimodal backbones to jointly perform perception, language conditioning, and action generation at every control step. Much of this capacity supports open-domain semantics, whereas continuous robot manipulation primarily requires compact representations of observations, actions, and the transitions induced by actions. Pixel-level world models provide another route, but predicting visual details irrelevant to control can be unnecessarily expensive. We propose SLIM (Self-supervised Latent Interaction Model), a compact 0.5B-parameter latent interaction policy. SLIM learns action-grounded predictive latents that capture both action-conditioned future transitions and the actions that explain observed changes. SLIM learns these representations through self-supervised masked trajectory prediction, combining action reconstruction with future-latent prediction. A compact Mixture-of-Transformers (MoT) backbone models interactions between observation latents and action tokens. The resulting policy is trained with flow matching for language-conditioned action generation. Across simulation benchmarks and real-world evaluation, SLIM matches or exceeds representative large-scale VLA and world-action-model baselines with fewer parameters, no additional embodied pretraining, lower inference latency, and substantially lower GPU memory usage.

cs.RO↗

arXiv2026

Tip-Tuned Renormalization-Group Spectroscopy Unmasks a False-positive Topological Superconducting Vortex

Clean, nonsplit vortex zero-bias peaks (ZBPs) can be misinterpreted as Majorana zero modes (MZMs), making static scanning tunneling microscopy intrinsically ambiguous. Here we use the STM tip coupling to drive a local boundary-renormalization-group (boundary RG) flow, turning dynamical Coulomb blockade into a falsification test for Majorana-like ZBPs. Experimentally, in a $\mathrm{SrSn}_3$ thin film, normal-state spectra establish an Ohmic dissipative environment, and a common boundary-RG/thermodynamic-Bethe-ansatz analysis of the superconducting-gap and vortex-center spectra yields consistent dissipation strengths within the $r < 1/2$ Majorana-filter regime. Lowering the tip nevertheless drives a clean, non-split vortex-center ZBP into a zero-bias dip, opposite to the protected flow of an isolated MZM, unmasking the peak as a Majorana false positive produced by a conventional vortex-core state. The same flow selectively suppresses the strongly tip-coupled channel, resolving the two-gap superconductivity. Dissipative STM thus tests dynamical protection rather than spectral appearance.

cond-mat.mes-hall↗

arXiv2026

UniDex: A Robot Foundation Suite for Universal Dexterous Hand Control from Egocentric Human Videos

Dexterous manipulation remains challenging due to the cost of collecting real-robot teleoperation data, the heterogeneity of hand embodiments, and the high dimensionality of control. We present UniDex, a robot foundation suite that couples a large-scale robot-centric dataset with a unified vision-language-action (VLA) policy and a practical human-data capture setup for universal dexterous hand control. First, we construct UniDex-Dataset, a robot-centric dataset over 50K trajectories across eight dexterous hands (6--24 DoFs), derived from egocentric human video datasets. To transform human data into robot-executable trajectories, we employ a human-in-the-loop retargeting procedure to align fingertip trajectories while preserving plausible hand-object contacts, and we operate on explicit 3D pointclouds with human hands masked to narrow kinematic and visual gaps. Second, we introduce the Function-Actuator-Aligned Space (FAAS), a unified action space that maps functionally similar actuators to shared coordinates, enabling cross-hand transfer. Leveraging FAAS as the action parameterization, we train UniDex-VLA, a 3D VLA policy pretrained on UniDex-Dataset and finetuned with task demonstrations. In addition, we build UniDex-Cap, a simple portable capture setup that records synchronized RGB-D streams and human hand poses and converts them into robot-executable trajectories to enable human-robot data co-training that reduces reliance on costly robot demonstrations. On challenging tool-use tasks across two different hands, UniDex-VLA achieves 81% average task progress and outperforms prior VLA baselines by a large margin, while exhibiting strong spatial, object, and zero-shot cross-hand generalization. Together, UniDex-Dataset, UniDex-VLA, and UniDex-Cap provide a scalable foundation suite for universal dexterous manipulation.

cs.RO↗

arXiv2026

Role of scaling dimensions in generalized noises in fractional quantum Hall tunneling due to a temperature bias

Continued improvement of heat control in mesoscopic conductors brings novel tools for probing strongly correlated electron phenomena. Motivated by these advances, we comprehensively study transport due to a temperature bias in a quantum point contact device in the fractional quantum Hall regime. We compute the charge-current noise (so-called delta-$T$ noise), heat-current noise, and mixed noise and elucidate how these observables can be used to infer strongly correlated properties of the device. Our main focus is the extraction of so-called scaling dimensions of the tunneling anyonic quasiparticles, of critical importance to correctly infer their anyonic exchange statistics.

cond-mat.mes-hall↗

arXiv2026

X-Distill: Cross-Architecture Vision Distillation for Visuomotor Learning

Visuomotor policies often leverage large pre-trained Vision Transformers (ViTs) for their powerful generalization capabilities. However, their significant data requirements present a major challenge in the data-scarce context of most robotic learning settings, where compact CNNs with strong inductive biases can be more easily optimized. To address this trade-off, we introduce X-Distill, a simple yet highly effective method that synergizes the strengths of both architectures. Our approach involves an offline, cross-architecture knowledge distillation, transferring the rich visual representations of a large, frozen DINOv2 teacher to a compact ResNet-18 student on the general-purpose ImageNet dataset. This distilled encoder, now endowed with powerful visual priors, is then jointly fine-tuned with a diffusion policy head on the target manipulation tasks. Extensive experiments on $34$ simulated benchmarks and $5$ challenging real-world tasks demonstrate that our method consistently outperforms policies equipped with from-scratch ResNet or fine-tuned DINOv2 encoders. Notably, X-Distill also surpasses 3D encoders that utilize privileged point cloud observations or much larger Vision-Language Models. Our work highlights the efficacy of a simple, well-founded distillation strategy for achieving state-of-the-art performance in data-efficient robotic manipulation.

cs.CV↗

arXiv2025

Detecting entanglement with transport measurement in weakly interacting and fluctuating systems

Measuring entanglement entropy in interacting, multipartite systems remains a significant experimental challenge. We address this challenge by developing a protocol to measure von Neumann entropy (VNE) and mutual information in quantum transport systems with both many-body interactions and multiple subsystems. Our analysis indicates that the vital connection between VNE and two-point correlation functions persists under these realistic conditions. The measurement is shown to be feasible for systems with boundary interactions and, critically, for bulk-interacting systems subject to a quantum quench of their internal couplings. Our work provides a pathway to experimentally quantify entanglement in complex interacting systems and establishes mutual information as an experimentally accessible indicator for system-environment entanglement.

cond-mat.mes-hall↗

arXiv2025

Transparent Graphene-Superconductor Interfaces: Quantum Hall and Zero Field Regimes

We study clean, edge-contacted graphene/superconductor interfaces in both the quantum Hall (QH) and zero field regimes. We find that Andreev reflection is substantially stronger than at an interface with a semiconductor two-dimensional electron gas: the large velocity at graphene's conical Dirac points makes the requirement of current continuity to a metal much less restrictive. In both our tight-binding and continuum models, we find a wide range of parameters for which Andreev reflection is strong. For a transparent interface, we demonstrate the following for graphene in the lowest Landau level QH state: (i) Excellent electron-hole hybridization occurs: the electron and hole components in graphene are simply related by an exchange of sublattice. The spatial profile for the electron component is predominantly gaussian on one sublattice and peaked at the interface, and so very different from the QH edge state of a terminated lattice. (ii) The degree of hybridization is independent of the graphene filling: no fine-tuning is needed. (iii) The spectrum is valley degenerate: the dispersion of each chiral Andreev edge mode (CAEM) self-aligns to be antisymmetric about the center of each valley, independent of filling. Achieving a transparent interface requires the absence of any barrier as well as a superconductor that is suitably matched to graphene; we argue that the latter condition is not very stringent. We further consider the effect of reduced transparency and Zeeman splitting on both the wavefunctions of the CAEM and their dispersion.

cond-mat.mes-hall↗

arXiv2025

Fractional-statistics-induced entanglement from Andreev-like tunneling

The role of anyonic statistics stands as a cornerstone in the landscape of topological quantum techniques. While recent years have brought forth encouraging and persuasive strides in detecting anyons, a significant facet remains unexplored, especially in view of connecting anyonic physics to quantum information platforms -- whether and how entanglement can be generated by anyonic braiding. Here, we demonstrate that even when two anyonic subsystems (represented by anyonic beams) are connected only by electron tunneling, entanglement between them, manifesting fractional statistics, is generated. To demonstrate this physics, we rely on a platform where fractional quantum Hall edges are bridged by a quantum point contact that allows only transmission of fermions (so-called Andreev-like tunneling). This invokes the physics of two-beam collisions in an anyonic Hong-Ou-Mandel collider, accompanied by a process that we dub anyon-quasihole braiding. We define an entanglement pointer -- a current-noise-based function tailored to quantify entanglement associated with quasiparticle fractional statistics. Our work, which exposes, both in theory and in experiment, entanglement associated with anyonic statistics and braiding, prospectively paves the way to the exploration of entanglement induced by non-Abelian statistics.

cond-mat.mes-hall↗

arXiv2025

Reactive Diffusion Policy: Slow-Fast Visual-Tactile Policy Learning for Contact-Rich Manipulation

Humans can accomplish complex contact-rich tasks using vision and touch, with highly reactive capabilities such as fast response to external changes and adaptive control of contact forces; however, this remains challenging for robots. Existing visual imitation learning (IL) approaches rely on action chunking to model complex behaviors, which lacks the ability to respond instantly to real-time tactile feedback during the chunk execution. Furthermore, most teleoperation systems struggle to provide fine-grained tactile / force feedback, which limits the range of tasks that can be performed. To address these challenges, we introduce TactAR, a low-cost teleoperation system that provides real-time tactile feedback through Augmented Reality (AR), along with Reactive Diffusion Policy (RDP), a novel slow-fast visual-tactile imitation learning algorithm for learning contact-rich manipulation skills. RDP employs a two-level hierarchy: (1) a slow latent diffusion policy for predicting high-level action chunks in latent space at low frequency, (2) a fast asymmetric tokenizer for closed-loop tactile feedback control at high frequency. This design enables both complex trajectory modeling and quick reactive behavior within a unified framework. Through extensive evaluation across three challenging contact-rich tasks, RDP significantly improves performance compared to state-of-the-art visual IL baselines. Furthermore, experiments show that RDP is applicable across different tactile / force sensors. Code and videos are available on https://reactive-diffusion-policy.github.io.

cs.RO↗

arXiv2025

ArrayBot: Reinforcement Learning for Generalizable Distributed Manipulation through Touch

We present ArrayBot, a distributed manipulation system consisting of a $16 \times 16$ array of vertically sliding pillars integrated with tactile sensors, which can simultaneously support, perceive, and manipulate the tabletop objects. Towards generalizable distributed manipulation, we leverage reinforcement learning (RL) algorithms for the automatic discovery of control policies. In the face of the massively redundant actions, we propose to reshape the action space by considering the spatially local action patch and the low-frequency actions in the frequency domain. With this reshaped action space, we train RL agents that can relocate diverse objects through tactile observations only. Surprisingly, we find that the discovered policy can not only generalize to unseen object shapes in the simulator but also transfer to the physical robot without any domain randomization. Leveraging the deployed policy, we present abundant real-world manipulation tasks, illustrating the vast potential of RL on ArrayBot for distributed manipulation.

cs.RO↗

arXiv2024

Landscapes of an out-of-equilibrium anyonic sea

The low-energy dynamics of two-dimensional topological matter hinges on its one-dimensional edge modes. Tunneling between fractional quantum Hall edge modes facilitates the study of anyonic statistics: it induces time-domain braiding that dominates signals from diluted anyon beams. We develop a framework for characterizing one-dimensional out-of-equilibrium anyonic states and define their effective potential and temperature, both arising from anyonic braiding, as well as the landscape of their excitations. Unlike fermions, the effective anyon potential depends on the type of the tunneling quasiparticles; non-equilibrium anyonic states are underlain by power-law energy distributions. This allows "hot" anyons to tunnel above the chemical potential of the source, which we capture by a measurable universal witness function. Our analysis raises the prospect of generalizing the kinetic approach to compressible anyonic matter in higher dimensions.

cond-mat.mes-hall↗

arXiv2024

Learning to Manipulate Anywhere: A Visual Generalizable Framework For Reinforcement Learning

Can we endow visuomotor robots with generalization capabilities to operate in diverse open-world scenarios? In this paper, we propose \textbf{Maniwhere}, a generalizable framework tailored for visual reinforcement learning, enabling the trained robot policies to generalize across a combination of multiple visual disturbance types. Specifically, we introduce a multi-view representation learning approach fused with Spatial Transformer Network (STN) module to capture shared semantic information and correspondences among different viewpoints. In addition, we employ a curriculum-based randomization and augmentation approach to stabilize the RL training process and strengthen the visual generalization ability. To exhibit the effectiveness of Maniwhere, we meticulously design 8 tasks encompassing articulate objects, bi-manual, and dexterous hand manipulation tasks, demonstrating Maniwhere's strong visual generalization and sim2real transfer abilities across 3 hardware platforms. Our experiments show that Maniwhere significantly outperforms existing state-of-the-art methods. Videos are provided at https://gemcollector.github.io/maniwhere/.

cs.RO↗

arXiv2024

Tunneling current and current correlations for anyonic quasiparticles of ν = 1/2 chiral Luttinger liquid in multi-edge geometries

We consider anyonic quasiparticles with charge e/2 described by the ν = 1/2 chiral Luttinger liquid, which collide in a Hong-Ou-Mandel-like interferometer. These colliding anyonic channels can be formally viewed as hosting Laughlin-like fractional ν = 1/2 quasiparticles. More specifically, two possible geometries are considered: (i) a two-edge-channel setup where anyons originate from equilibrium reservoirs; (ii) a four-edge-channel setup where nonequilibrium anyons arrive at the collider in the form of diluted beams. For both setups, we calculate the tunneling current and the current correlations. For setup (i), our results provide analytically exact expressions for the tunneling current, tunneling-current noise, and cross-correlation noise, The exact relation between conductance and noise is demonstrated. For setup (ii), we show that the tunneling current and the generalized Fano factor [defined in B. Rosenow et al. (2016)] are finite for diluted streams of ν = 1/2 anyons. This is due to the processes where nonequilibrium anyons, supplied via either source edge, directly tunnel at the central QPC. Thus, to obtain meaningful results in this case, one should go beyond the so-called time-domain braiding processes, where nonequilibrium anyons do not tunnel at the collider, but rather indirectly influence the tunneling by braiding with the quasiparticle-quasihole pairs created at the collider. This suggests that the effect of direct tunneling and collisions of diluted anyons in the Hong-Ou-Mandel interferometer can be important for various observables in physical quantum-Hall edges at Laughlin filling fractions.

cond-mat.mes-hall↗

arXiv2024

3D Diffusion Policy: Generalizable Visuomotor Policy Learning via Simple 3D Representations

Imitation learning provides an efficient way to teach robots dexterous skills; however, learning complex skills robustly and generalizablely usually consumes large amounts of human demonstrations. To tackle this challenging problem, we present 3D Diffusion Policy (DP3), a novel visual imitation learning approach that incorporates the power of 3D visual representations into diffusion policies, a class of conditional action generative models. The core design of DP3 is the utilization of a compact 3D visual representation, extracted from sparse point clouds with an efficient point encoder. In our experiments involving 72 simulation tasks, DP3 successfully handles most tasks with just 10 demonstrations and surpasses baselines with a 24.2% relative improvement. In 4 real robot tasks, DP3 demonstrates precise control with a high success rate of 85%, given only 40 demonstrations of each task, and shows excellent generalization abilities in diverse aspects, including space, viewpoint, appearance, and instance. Interestingly, in real robot experiments, DP3 rarely violates safety requirements, in contrast to baseline methods which frequently do, necessitating human intervention. Our extensive evaluation highlights the critical importance of 3D representations in real-world robot learning. Videos, code, and data are available on https://3d-diffusion-policy.github.io .

cs.RO↗

arXiv2024

Influence of topological degeneracy on the boundary Berezinskii-Kosterlitz-Thouless quantum phase transition of a dissipative resonant level

The interplay between a topological degeneracy and the residue degeneracy (also known as the residue entropy) of quantum criticality remains as an important but not thoroughly understood topic. We find that this topological degeneracy, provided by a Majorana zero mode pair, relaxes the otherwise strictly requested symmetry requirement, to observe the boundary Berezinskii-Kosterlitz-Thouless (BKT) quantum phase transition (QPT) of a dissipative resonant level. Our work indicates that the topological degeneracy can be potentially viewed as an auxiliary symmetry that realizes a robust boundary QPT. The relaxation of the symmetry requirement extends the transition from a point to a finite area, thus greatly reducing the difficulty to experimentally observe the QPT. This topology-involved exotic BKT phase diagram, on the other hand, provides another piece of evidence that can further confirm the existence of a Majorana zero mode.

cond-mat.mes-hall↗

arXiv2024

Thin-Shell Object Manipulations With Differentiable Physics Simulations

In this work, we aim to teach robots to manipulate various thin-shell materials. Prior works studying thin-shell object manipulation mostly rely on heuristic policies or learn policies from real-world video demonstrations, and only focus on limited material types and tasks (e.g., cloth unfolding). However, these approaches face significant challenges when extended to a wider variety of thin-shell materials and a diverse range of tasks. While virtual simulations are shown to be effective in diverse robot skill learning and evaluation, prior thin-shell simulation environments only support a subset of thin-shell materials, which also limits their supported range of tasks. We introduce ThinShellLab - a fully differentiable simulation platform tailored for robotic interactions with diverse thin-shell materials possessing varying material properties, enabling flexible thin-shell manipulation skill learning and evaluation. Our experiments suggest that manipulating thin-shell objects presents several unique challenges: 1) thin-shell manipulation relies heavily on frictional forces due to the objects' co-dimensional nature, 2) the materials being manipulated are highly sensitive to minimal variations in interaction actions, and 3) the constant and frequent alteration in contact pairs makes trajectory optimization methods susceptible to local optima, and neither standard reinforcement learning algorithms nor trajectory optimization methods (either gradient-based or gradient-free) are able to solve the tasks alone. To overcome these challenges, we present an optimization scheme that couples sampling-based trajectory optimization and gradient-based optimization, boosting both learning efficiency and converged performance across various proposed tasks. In addition, the differentiable nature of our platform facilitates a smooth sim-to-real transition.

cs.RO↗