arXiv subjects

Zewen Wu

Publications and source records attributed to Zewen Wu.

18 recordsLinked to original sources

arXiv2026

An AI-driven robotic system for two-dimensional hetero-assemblies

Nanomaterials stacked on-demand, such as rotationally assembled two-dimensional (2D) van der Waals (vdW) layered compounds, provides a versatile platform for quantum simulation and the exploration of exotic electronic phases. Currently, however, such nanoassemblies remain largely confined to inefficiency, manually operated process, limiting their potential for probing emergent physical phenomena. There is a pressing need in the field for high-precision, automated assembling techniques, especially for the scalable fabrication of 2D twistronic heterostructures. Here, we present an intelligent automation system dedicated to the fabrication of van der Waals stacks, following the state-of-the-art protocol for dry transfer of exfoliated 2D materials. The system further employs metadata generated from each automated stacking procedure to perform reinforcement learning, thereby continuously bettering its performances. As a concrete demonstration, we fabricate twisted bilayer graphene (TBLG) -- known for its challenging preparation -- and exhibit its unconventional superconductivity near the magic angle. Our work may pave the way for high-throughput fabrication of low-dimensional nanomaterials including twistronic heterostructures, where integrating data mining and artificial intelligence can accelerate the discovery of novel physical phenomena.

cond-mat.mes-hall↗

arXiv2026

Electromechanical Switching and Momentum-Selective Transport in Geometry-Defined Blue Phosphorus Homojunctions

Developing intrinsic homojunctions without chemical heterogeneity remains a key challenge in future two - dimensional devices. Here, we report a geometry - defined metal--semiconductor--metal homojunction in bilayer blue phosphorus (BlueP) created by a localized bubble corrugation, without chemical doping or foreign - material interfaces. First - principles calculations show that enlarging the interlayer separation in the metallic A$_1$B$_1$ - stacked BlueP bilayer opens a band gap, enabling a semiconducting barrier embedded between metallic segments. First - principles quantum - transport simulations reveal a crossover from ballistic to tunneling transport upon bubble formation. In the tunneling regime, transmission decreases exponentially with bubble width while remaining weakly sensitive to bubble height and bulging direction. The junction acts as an orientation - dependent $k$ - space filter, producing transport anisotropy and momentum selectivity. Orbital - resolved scattering analysis shows that intralayer - bonding channels persist under deformation whereas interlayer - hybridized channels are quenched, and that \sigma - type bonding yields higher conductance than \pi - type bonding. These insights motivate two electromechanical device concepts: a mechanically switchable memory element with ON/OFF ratios up to 30 and a nanoscale sliding rheostat with reproducible exponential resistance tuning for \AA - scale displacement sensing.

cond-mat.mes-hall↗

arXiv2025

GenDexHand: Generative Simulation for Dexterous Hands

Data scarcity remains a fundamental bottleneck for embodied intelligence. Existing approaches use large language models (LLMs) to automate gripper-based simulation generation, but they transfer poorly to dexterous manipulation, which demands more specialized environment design. Meanwhile, dexterous manipulation tasks are inherently more difficult due to their higher degrees of freedom. Massively generating feasible and trainable dexterous hand tasks remains an open challenge. To this end, we present GenDexHand, a generative simulation pipeline that autonomously produces diverse robotic tasks and environments for dexterous manipulation. GenDexHand introduces a closed-loop refinement process that adjusts object placements and scales based on vision-language model (VLM) feedback, substantially improving the average quality of generated environments. Each task is further decomposed into sub-tasks to enable sequential reinforcement learning, reducing training time and increasing success rates. Our work provides a viable path toward scalable training of diverse dexterous hand behaviors in embodied intelligence by offering a simulation-based solution to synthetic data generation. Our website: https://winniechen2002.github.io/GenDexHand/.

cs.RO↗

arXiv2025

TBPLaS 2.0: a Tight-Binding Package for Large-scale Simulation

The common exact diagonalization-based techniques to solving tight-binding models suffer from O(N^2) and O(N^3) scaling with respect to model size in memory and CPU time, hindering their applications in large tight-binding models. On the contrary, the tight-binding propagation method (TBPM) can achieve linear scaling in both memory and CPU time, and is capable of handling large tight-binding models with billions of orbitals. In this paper, we introduce version 2.0 of TBPLaS, a package for large-scale simulation based on TBPM. This new version brings significant improvements with many new features. Existing Python/Cython modeling tools have been thoroughly optimized, and a compatible C++ implementation of the modeling tools is now available, offering efficiency enhancement of several orders. The solvers have been rewritten in C++ from scratch, with the efficiency enhanced by several times or even by an order of magnitude. The workflow of utilizing solvers has also been unified into a more comprehensive and consistent manner. New features include spin texture, Berry curvature and Chern number calculation, partial diagonalization for specific eigenvalues and eigenstates, analytical Hamiltonian, and GPU computing support. The documentation and tutorials have also been updated to the new version. In this paper, we discuss the revisions with respect to version 1.3 and demonstrate the new features. Benchmarks on modeling tools and solvers are also provided.

physics.comp-ph↗

arXiv2025

Semiregular tessellation of electronic lattices in untwisted bilayer graphene under anisotropic strain gradients

Two-dimensional (2D) moir\'e superlattices have emerged as a versatile platform for uncovering exotic quantum phases, many of which arise in bilayer systems exhibiting Archimedean tessellation patterns such as triangular, hexagonal, and kagome lattices. Here, we propose a strategy to engineer semiregular tessellation patterns in untwisted bilayer graphene by applying anisotropic epitaxial tensile strain (AETS) along crystallographic directions. Through force-field and first-principles calculations, we demonstrate that AETS can induce a rich variety of semiregular tessellation geometries, including truncated hextille, prismatic pentagon, and brick-phase arrangements. The characteristic electronic bands (Dirac and flat bands) of the lattice models associated with these semiregular tessellations are observed near the Fermi level, arising from interlayer interactions generated by the redistribution of specific stacking registries (AB, BA, and SP). Furthermore, the electronic kagome, distorted Lieb, brick-like, and one-dimensional stripe lattices captured in real-space confirm the tunable nature of the semiregular tessellation lattices enabled by AETS. Our study identifies AETS as a promising new degree of freedom in moir\'e engineering, offering a reproducible and scalable platform for exploring exotic electronic lattices in moir\'e systems.

cond-mat.mes-hall↗

arXiv2025

Scaling Up On-Device LLMs via Active-Weight Swapping Between DRAM and Flash

Large language models (LLMs) are increasingly being deployed on mobile devices, but the limited DRAM capacity constrains the deployable model size. This paper introduces ActiveFlow, the first LLM inference framework that can achieve adaptive DRAM usage for modern LLMs (not ReLU-based), enabling the scaling up of deployable model sizes. The framework is based on the novel concept of active weight DRAM-flash swapping and incorporates three novel techniques: (1) Cross-layer active weights preloading. It uses the activations from the current layer to predict the active weights of several subsequent layers, enabling computation and data loading to overlap, as well as facilitating large I/O transfers. (2) Sparsity-aware self-distillation. It adjusts the active weights to align with the dense-model output distribution, compensating for approximations introduced by contextual sparsity. (3) Active weight DRAM-flash swapping pipeline. It orchestrates the DRAM space allocation among the hot weight cache, preloaded active weights, and computation-involved weights based on available memory. Results show ActiveFlow achieves the performance-cost Pareto frontier compared to existing efficiency optimization methods.

cs.LG↗

arXiv2024

Surface region band enhancement in noble gas adsorption assisted ARPES on kagome superconductor RbV3Sb5

Electronic states near surface regions can be distinct from bulk states, which are paramount in understanding various physical phenomena occurring at surfaces and in applications in semiconductors, energy, and catalysis. Here, we report an abnormal surface region band enhancement effect in angle-resolved photoemission spectroscopy on kagome superconductor RbV3Sb5, by depositing noble gases with fine control. In contrast to conventional surface contamination, the intensity of surface region Sb band can be enhanced more than three times with noble gas adsorption. In the meantime, a hole-dope effect is observed for the enhanced surface region band, with other bands hardly changing. The doping effect is more pronounced with heavier noble gases. We propose that noble gas atoms selectively fill into alkali metal vacancy sites on the surface, which improves the surface condition, boosts surface region bands, and effectively dopes it with the Pauli repulsion mechanism. Our results provide a novel and reversible way to improve surface conditions and tune surface region bands by controlled surface noble gas deposition.

cond-mat.mtrl-sci↗

arXiv2023

Kagome electronic states in gradient-strained untwisted graphene bilayers

Moir\'e superlattices in twisted homo-bilayers have revealed exotic electronic states, including unconventional superconductivity and correlated insulating phases. However, their fabrication process often introduces moir\'e disorders, hindering reproducibility and experimental control. Here, we propose an alternative approach using gradient strain to construct moir\'e superlattices in untwisted bilayer graphene (gs-BLG). Through force-field and first-principles calculations, we show that gs-BLG exhibits kagome-like interlayer-spacing distributions and strain-tunable kagome electronic bands. The competition between interlayer coupling and in-plane strain relaxation leads to distinct structural deformations, giving rise to three forms of diatomic kagome lattices: subtle, pronounced, and distorted. Kagome electronic bands are identified near the Fermi level in their band structures. Modulating strain gradients enables tailoring bandwidths and signs of hopping parameters of these kagome bands, providing a versatile platform for studying exotic electronic phases. Our findings establish gradient strain as an alternative to twist engineering, opening an avenue for exploring emergent electronic phases in graphene-based systems.

cond-mat.mes-hall↗

arXiv2023

Post-training Quantization for Text-to-Image Diffusion Models with Progressive Calibration and Activation Relaxing

High computational overhead is a troublesome problem for diffusion models. Recent studies have leveraged post-training quantization (PTQ) to compress diffusion models. However, most of them only focus on unconditional models, leaving the quantization of widely-used pretrained text-to-image models, e.g., Stable Diffusion, largely unexplored. In this paper, we propose a novel post-training quantization method PCR (Progressive Calibration and Relaxing) for text-to-image diffusion models, which consists of a progressive calibration strategy that considers the accumulated quantization error across timesteps, and an activation relaxing strategy that improves the performance with negligible cost. Additionally, we demonstrate the previous metrics for text-to-image diffusion model quantization are not accurate due to the distribution gap. To tackle the problem, we propose a novel QDiffBench benchmark, which utilizes data in the same domain for more accurate evaluation. Besides, QDiffBench also considers the generalization performance of the quantized model outside the calibration dataset. Extensive experiments on Stable Diffusion and Stable Diffusion XL demonstrate the superiority of our method and benchmark. Moreover, we are the first to achieve quantization for Stable Diffusion XL while maintaining the performance.

cs.CV↗

arXiv2023

Prioritized Planning for Target-Oriented Manipulation via Hierarchical Stacking Relationship Prediction

In scenarios involving the grasping of multiple targets, the learning of stacking relationships between objects is fundamental for robots to execute safely and efficiently. However, current methods lack subdivision for the hierarchy of stacking relationship types. In scenes where objects are mostly stacked in an orderly manner, they are incapable of performing human-like and high-efficient grasping decisions. This paper proposes a perception-planning method to distinguish different stacking types between objects and generate prioritized manipulation order decisions based on given target designations. We utilize a Hierarchical Stacking Relationship Network (HSRN) to discriminate the hierarchy of stacking and generate a refined Stacking Relationship Tree (SRT) for relationship description. Considering that objects with high stacking stability can be grasped together if necessary, we introduce an elaborate decision-making planner based on the Partially Observable Markov Decision Process (POMDP), which leverages observations and generates the least grasp-consuming decision chain with robustness and is suitable for simultaneously specifying multiple targets. To verify our work, we set the scene to the dining table and augment the REGRAD dataset with a set of common tableware models for network training. Experiments show that our method effectively generates grasping decisions that conform to human requirements, and improves the implementation efficiency compared with existing methods on the basis of guaranteeing the success rate.

cs.RO↗

arXiv2021

Realizing One-dimensional Metallic States in Graphene via Periodically Coupled Zeroth Pseudo-Landau Levels

Strain-induced pseudo-magnetic fields can mimic real magnetic fields to generate a zero-magnetic-field analogue of the Landau levels (LLs), i.e., the pseudo-LLs, in graphene. The distinct nature of the pseudo-LLs enables one to realize novel electronic states beyond that can be feasible with real LLs. Here, we report the realization of one-dimensional (1D) metallic states, which can be described well by the Su-Schrieffer-Heeger model, in graphene via periodically coupled zeroth pseudo-LLs. In our experiment, nanoscale strained structures embedded with pseudo-LLs are generated periodically along 1D channel of suspended graphene monolayer. Our experiments demonstrate that the zeroth pseudo-LLs of these strained structures are coupled to form metallic states, exhibiting a serpentine pattern that snakes back and forth along the 1D suspended graphene monolayer. These results are verified theoretically by large-scale tight-binding calculations of the strained samples. Our result provides a new pathway to realize novel quantum states and engineer the electronic properties of graphene by using the localized pseudo-LLs as building blocks.

cond-mat.mes-hall↗

arXiv2021

Magic angle and plasmon mode engineering in twisted trilayer graphene with pressure

Recent experimental and theoretical investigations demonstrate that twisted trilayer graphene (tTLG) is a highly tunable platform to study the correlated insulating states, ferromagnetism, and superconducting properties. Here we explore the possibility of tuning electronic correlations of the tTLG via a vertical pressure. A full tight-binding model is used to accurately describe the pressure-dependent interlayer interactions. Our results show that pressure can push a relatively larger twist angle (for instance, $1.89^{\circ}$) tTLG to reach the flat-band regime. Next, we obtain the relationship between the pressure-induced magic angle value and the critical pressure. These critical pressure values are almost half of that needed in the case of twisted bilayer graphene. Then, plasmonic properties are further investigated in the flat band tTLG with both zero-pressure magic angle and pressure-induced magic angle. Two plasmonic modes are detected in these two kinds of flat band samples. By comparison, one is a high energy damping-free plasmon mode that shows similar behavior, and the other is a low energy plasmon mode (flat-band plasmon) that shows obvious differences. The flat-band plasmon is contributed by both interband and intraband transitions of flat bands, and its divergence is due to the various shape of the flat bands in tTLG with zero-pressure and pressure-induced magic angles. This may provide an efficient way of tuning between regimes with strong and weak electronic interactions in one sample and overcoming the technical requirement of precise control of the twist angle in the study of correlated physics.

cond-mat.mes-hall↗

arXiv2021

Lattice relaxation, mirror symmetry and magnetic field effects on ultraflat bands in twisted trilayer graphene

Twisted graphene multilayers exhibit strongly correlated insulating states and superconductivity due to the presence of ultraflat bands near the charge neutral point. In this paper, the response of ultraflat bands to lattice relaxation and a magnetic field in twisted trilayer graphene (tTLG) with different stacking arrangements is investigated by using a full tight-binding model. We show that lattice relaxations are indispensable for understanding the electronic properties of tTLG, in particular, of tTLG in the presence of mirror symmetry. Lattice relaxations renormalize the quasiparticle spectrum near the Fermi energy and change the localization of higher energy flat bands. Furthermore, different from the twisted bilayer graphene, the Hofstadter butterfly spectrum can be realized at laboratory accessible strengths of magnetic field. Our work verifies tTLG as a more tunable platform than the twisted bilayer graphene in strongly correlated phenomena.

cond-mat.mes-hall↗

arXiv2020

Electronic structure of 30° twisted double bilayer graphene

In this paper, the electronic properties of 30° twisted double bilayer graphene, which loses the translational symmetry due to the incommensurate twist angle, are studied by means of the tight-binding approximation. We demonstrate the interlayer decoupling in the low-energy region from various electronic properties, such as the density of states, effective band structure, optical conductivity and Landau level spectrum. However, at Q points, the interlayer coupling results in the appearance of new Van Hove singularities in the density of states, new peaks in the optical conductivity and importantly the 12-fold-symmetry-like electronic states. The k-space tight-binding method is adopted to explain this phenomenon. The electronic states at Q points show the charge distribution patterns more complex than the 30° twisted bilayer graphene due to the symmetry decrease. These phenomena appear also in the 30° twisted interface between graphene monolayer and AB stacked bilayer.

cond-mat.mtrl-sci↗

arXiv2019

Dodecagonal bilayer graphene quasicrystal and its approximants

Dodecagonal bilayer graphene quasicrystal has 12-fold rotational order but lacks translational symmetry which prevents the application of band theory. In this paper, we study the electronic and optical properties of graphene quasicrystal with large-scale tight-binding calculations involving more than ten million atoms. We propose a series of periodic approximants which reproduce accurately the properties of quasicrystal within a finite unit cell. By utilizing the band-unfolding method on the smallest approximant with only 2702 atoms, the effective band structure of graphene quasicrystal is derived. Novel features, such as the emergence of new Dirac points (especially the mirrored ones), the band gap at M point and the Fermi velocity are all in agreement with recent experiments. The properties of quasicrystal states are identified in the Landau level spectrum and optical excitations. Importantly, our results show that the lattice mismatch is the dominant factor determining the accuracy of layered approximants. The proposed approximants can be used directly for other layered materials in honeycomb lattice, and the design principles can be applied for any quasi-periodic incommensurate structures.

cond-mat.mes-hall↗

arXiv2019

Strain-tunable magnetic and electronic properties of monolayer CrI3

Two-dimensional CrI3 has attracted much attention as it is reported to be a ferromagnetic semiconductor with the Curie temperature around 45K. By performing first-principles calculations, we find that the magnetic ground state of CrI3 is variable under biaxial strain. Our theoretical investigations show that the ground state of monolayer CrI3 is ferromagnetic under compression, but becomes antiferromagnetic under tension. Particularly, the transition occurs under a feasible in-plane strain around 1.8%. Accompanied by the transition of the magnetic ground state, it undergoes a transition from magnetic-metal to half-metal to half-semiconductor to spin-relevant semiconductor when strain varies from -15% to 10%. We attribute these transitions to the variation of the d-orbitals of Cr atoms and the p-orbitals of I atoms. Generally, we report a series of magnetic and electronic phase transition in strained CrI3, which will help both theoretical and experimental researchers for further understanding of the tunable electronic and magnetic properties of CrI3 and their analogous.

physics.comp-ph↗

arXiv2018

Two-Dimensional Photogalvanic Spin-Battery

Pure spin-current is of central importance in spintronics. Here we propose a two-dimensional (2D) spin-battery system that delivers pure spin-current without an accompanying charge-current to the outside world at zero bias. The principle of the spin-battery roots in the photogalvanic effect (PGE), and the system has good operational stability against structural perturbation, photon energy and other materials detail. The device principle is numerically implemented in the 2D material phosphorene as an example, and first principles calculations give excellent qualitative agreement with the PGE physics. The 2D spin-battery is interesting as it is both a device that generates pure spin-currents, also an energy source that harvests photons. Given the versatile operational space, the spin-battery should be experimentally feasible.

physics.app-ph↗

arXiv2018

First principles investigation of pressure related quantum transport in pure black phosphorus devices

We propose a first-principles calculation to investigate the pressure-related transport properties of two kinds of pure monolayer black phosphorus (MBP) devices. Numerical results show that semi-conducting MBP can withstand a considerable compression pressure until it is transformed to be a conductor. The pure MBP devices can work as flexible electronic devices, "negative" pressure sensors, and "positive" pressure sensors depending on the chirality of BP and the magnitude of vertical pressure. When pressure is relatively small, the conductance is robust against the stress for zigzag MBP devices, while shows pressure-sensitive properties for armchair MBP devices. The pressure-stable property of zigzag MBP devices implies a good application prospects as flexible electronic devices, however, the distinct negative increase of conductance versus pressure indicates that armchair MBP devices can work as "negative" pressure sensors. When pressure is relatively large, both armchair MBP devices and zigzag MBP devices show favorable properties of "positive" pressure sensors, whose conductivities rise promptly versus pressure. The longer the device, the more the pressure sensitivity. Band alignment analysis and empirical Wentzel$-$Kramers$-$Brillouin (WKB) approximations are also performed to testify the tunneling process of pure MBP devices from first principles calculation.

cond-mat.mtrl-sci↗