arXiv subjects

Du Li

Publications and source records attributed to Du Li.

14 recordsLinked to original sources

arXiv2026

Anomalous stabilization of excitons by metallic proximity

Metallic environments are generally expected to suppress excitons through strong dielectric screening, yet their influence can differ in composite systems where metallic and insulating regions coexist. Here, we investigate excitonic states in PdxCu1-xCrO2 thin films across a percolation-driven metal-insulator transition. Optical spectroscopy and many-body GW calculations show that CuCrO2 hosts strongly bound excitons with a binding energy of about 489 meV. With increasing Pd substitution, the system approaches an insulator-to-metal transition near x = 0.5, consistent with the site-percolation threshold of a triangular lattice. In the pre-percolation regime, the excitonic resonance redshifts by 241 meV while the Tanguy continuum onset remains nearly unchanged, consistent with a substantial increase in exciton binding energy before metallization. An image-charge-based excitonic hydrogen model shows that isolated metallic regions can enhance electron-hole binding through image-charge interactions, whereas conventional screening is recovered once a continuous metallic network forms. Although this model provides a possible interpretation of the observed excitonic evolution, an alternative scenario in which metallic and excitonic responses originate from electronically distinct states and evolve independently cannot be excluded. These results reveal unusual metallic-excitonic coexistence near a percolation-driven metal-insulator transition and suggest nanoscale metallic proximity as a possible route for modifying excitonic interactions.

cond-mat.str-el↗

arXiv2026

Twist-induced Out-of-plane Ferroelectricity in Bilayer Hafnia

Ferroelectric HfO2 is a promising candidate for next-generation memory devices due to its CMOS compatibility and ability to retain polarization at nanometer scales. However, the polar orthorhombic phase (Pca2_1) responsible for ferroelectricity is metastable and requires extrinsic stabilization, which makes it challenging for integration with silicon. We predict that bilayer 1T-HfO2 can exhibit robust and switchable out-of-plane (OOP) polarization arising from stacking-induced symmetry breaking. Using first-principles density functional theory, we predict that monolayer 1T-HfO2 can be cleaved from the (111) surface of cubic hafnia, and the monolayer is dynamically stable. When two aligned monolayers are twisted to form a moir\'e superlattice, it breaks the interlayer symmetry and allows the emergence of bistable OOP polarization. At a twist angle of 7.34o, the system exhibits a net polarization of ~16 {\mu}C/cm2. This sizeable polarization is due to the large polar displacements concentrated in AB stacking domains. Importantly, this polarization can be reversibly switched via interlayer sliding with a low energy barrier (~8 meV/formula unit) and comparable low coercive field (~0.2 V/nm), offering electric-field tunability. These findings establish twisted bilayer 1T-HfO2 as a scalable and robust 2D ferroelectric platform, enabling new pathways for integrating ferroelectric functionality into atomically thin memory and logic devices.

cond-mat.mtrl-sci↗

arXiv2025

Virtual Width Networks

We introduce Virtual Width Networks (VWN), a framework that delivers the benefits of wider representations without incurring the quadratic cost of increasing the hidden size. VWN decouples representational width from backbone width, expanding the embedding space while keeping backbone compute nearly constant. In our large-scale experiment, an 8-times expansion accelerates optimization by over 2 times for next-token and 3 times for next-2-token prediction. The advantage amplifies over training as both the loss gap grows and the convergence-speedup ratio increases, showing that VWN is not only token-efficient but also increasingly effective with scale. Moreover, we identify an approximately log-linear scaling relation between virtual width and loss reduction, offering an initial empirical basis and motivation for exploring virtual-width scaling as a new dimension of large-model efficiency.

cs.LG↗

arXiv2025

Valley-Selective Linear Dichroism and Excitonic Effects in Lieb-Lattice Altermagnets

Altermagnets have recently been recognized as a distinct class of magnetic materials characterized by alternative spin-split electronic structures without net magnetization. Despite intensive studies on their single-particle spintronic and valleytronic properties, many-electron interactions and optical responses of altermagnets remain less explored. In this work, we employ many-body perturbation theory to investigate excited states and their strain tunability. Using monolayer Mn2WS4 as a representative candidate, we uncover a novel spin valley-dependent excitonic selection rule in two-dimensional altermagnetic Lieb lattices. In addition to strongly bound excitons, we find that linearly polarized light selectively excites valley spin-polarized excitons. Moreover, due to the interplay between altermagnetic spin symmetry and electronic orbital character, we predict that applying uniaxial strain can lift valley degeneracy and enable the selective excitation of spin-polarized excitons, an effect not achievable in previously studied transition-metal dichalcogenides. These spin-valley-locked excitonic states and their strain tunability offer a robust mechanism for four-fold symmetric altermagnets to encode, store, and read valley/spin information.

cond-mat.mtrl-sci↗

arXiv2025

Quasiparticle energies and excitonic effects of {\alpha}-RuCl3

{\alpha}-phase Ruthenium (III) chloride ({\alpha}-RuCl3) has attracted significant attention because of its potential for realizing Kitaev quantum spin liquid. In this work, we employ first-principles many-body perturbation theory to study its many-electron interactions and excited-state properties. We find enhanced many-electron interactions that dominate quasiparticle energies and optical responses in {\alpha}-RuCl3. Our calculated quasiparticle band gap of bulk structure is about 1.75 eV that agrees well with recent scanning tunneling spectroscopy and angle-resolved photoemission spectroscopy measurements. Furthermore, our calculated first and second primary bright excitons are located at 1.23 eV and 1.98 eV, respectively. These excitons show good consistency with the main features observed in the optical absorption spectrum. We extend our investigation to monolayer {\alpha}-RuCl3, examining the zigzag antiferromagnetic (AFM) and ferromagnetic (FM) phases. In addition to significant excitonic effects, the optical spectrum of the zigzag AFM phase exhibits anisotropic behavior, while the FM phase demonstrates isotropic characteristics. The different optical response behaviors provide an efficient approach to identify the energy nearly degenerate magnetic states, which can both potentially exist in fabricated samples.

cond-mat.mtrl-sci↗

arXiv2025

DeepCompile: A Compiler-Driven Approach to Optimizing Distributed Deep Learning Training

The rapid growth of deep learning models has increased the demand for efficient distributed training strategies. Fully sharded approaches like ZeRO-3 and FSDP partition model parameters across GPUs and apply optimizations such as prefetching and unsharding to reduce communication overhead. However, these systems lack fine-grained control over memory and communication scheduling, making it difficult to balance computation--communication overlap with memory requirements. Coordinating multiple optimizations such as prefetching and unsharding is also difficult, since their effects on memory usage can influence each other. To tackle these challenges, we propose DeepCompile, a compiler-based optimization framework for distributed training. DeepCompile transforms user-defined models into computation graphs and applies a series of profiling-guided optimization passes, each modifying the graph based on profiling information such as execution time and memory usage. This design allows each pass to flexibly insert, reorder, or remove operations such as all-gather and memory allocation, improving communication--computation overlap and reducing memory pressure. Each pass can access updated profiling feedback from earlier passes, enabling coordinated optimizations. We further enhance DeepCompile by three additional optimizations: proactive prefetching, selective unsharding, and adaptive offloading. Our evaluation shows that DeepCompile achieves up to 1.28$\times$ and 1.54$\times$ speedups over ZeRO-3 and FSDP baselines, respectively, and up to a 7.01$\times$ throughput increase in settings with limited GPU resources using offloading.

cs.DC↗

arXiv2025

Itinerant Magnetism in Twisted Bilayer WSe$_2$ and MoTe$_2$

Using a self-consistent Hartree-Fock theory, we show that the recently observed ferromagnetism in twisted bilayer WSe$_2$ [Nat. Commun. 16, 1959 (2025)] can be understood as a Stoner-like instability of interaction-renormalized moir\'e bands. We quantitatively reproduce the observed Lifshitz transition as function of hole filling and applied electric field that marks the boundary between layer-hybridized and layer-polarized regimes. The former supports a ferromagnetic valley-polarized ground state below half-filling, developing a topological charge gap at half-filling for smaller twist angles. At larger twist angles, the system hosts a gapped triangular N\'eel antiferromagnet. On the other hand, the layer-polarized regime supports a stripe antiferromagnet below half-filling and a wing-shaped multiferroic ground state above half-filling. We map the evolution of these states as a function of filling factor, electric field, twist angle, and interaction strength. Our results demonstrate that long-range exchange in a symmetry-unbroken parent state with strongly renormalized moir\'e bands provides a broadly applicable framework to understand itinerant magnetism in moir\'e TMDs.

cond-mat.str-el↗

arXiv2024

UOTe: Kondo-interacting topological antiferromagnet in a van der Waals lattice

Since the initial discovery of two-dimensional van der Waals (vdW) materials, significant effort has been made to incorporate the three properties of magnetism, band structure topology, and strong electron correlations $-$ to leverage emergent quantum phenomena and expand their potential applications. However, the discovery of a single vdW material that intrinsically hosts all three ingredients has remained an outstanding challenge. Here we report the discovery of a Kondo-interacting topological antiferromagnet in the vdW 5$f$ electron system UOTe. It has a high antiferromagnetic (AFM) transition temperature of 150 K, with a unique AFM configuration that breaks the combined parity and time reversal ($PT$) symmetry in an even number of layers while maintaining zero net magnetic moment. Our angle-resolved photoemission spectroscopy (ARPES) measurements reveal Dirac bands near the Fermi level, which combined with our theoretical calculations demonstrate UOTe as an AFM Dirac semimetal. Within the AFM order, we observed the presence of the Kondo interaction, as evidenced by the emergence of a 5$f$ flat band near the Fermi level below 100 K and hybridization between the Kondo band and the Dirac band. Our density functional theory calculations in its bilayer form predict UOTe as a rare example of a fully-compensated AFM Chern insulator.

cond-mat.str-el↗

arXiv2024

Insulator to Metal Transition under High Pressure in FeNb$_3$Se$_{10}$

Non-magnetic FeNb$_3$Se$_{10}$ has been demonstrated to be an insulator at ambient pressure through both theoretical calculations and experimental measurements and it does not host topological surface states. Here we show that on the application of pressure, FeNb$_3$Se$_{10}$ transitions to a metallic state at around 3.0 GPa. With a further increase in pressure, its resistivity becomes independent of both temperature and pressure. Its crystal structure is maintained to at least 4.4 GPa.

cond-mat.str-el↗

arXiv2023

Accelerating GW calculations of point defects with the defect-patched screening approximation

The GW approximation has been widely accepted as an ab initio tool for calculating defect levels with many-electron effect included. However, the GW simulation cost increases dramatically with the system size, and, unfortunately, large supercells are often required to model low-density defects that are experimentally relevant. In this work, we propose to accelerate GW calculations of point defects by reducing the simulation cost of the many-electron screening, which is the primary computational bottleneck. The random-phase approximation of many-electron screening is divided into two parts: one is the intrinsic screening, calculated using a unit cell of pristine structures, and the other is the defect-induced screening, calculated using the supercell within a small energy window. Depending on specific defects, one may only need to consider the intrinsic screening or include the defect contribution. This approach avoids the summation of many conductions states of supercells and significantly reduces the simulation time. We have applied it to calculating various point defects, including neutral and charged defects in two-dimensional and bulk systems with small or large bandgaps. The results consist with those from the direct GW simulations, and the agreements are further improved at the dilute-defect limit, which is experimentally relevant but extremely challenging for direct GW simulations. This defect-patched screening approach not only clarifies the roles of defects in many-electron screening but also paves the way to fast screen defect structures/materials for novel applications, including single-photon sources, quantum qubits, and quantum sensors.

cond-mat.mtrl-sci↗

arXiv2023

Differentiating $L_\infty$ groupoids

Differentiating an Lie $n$-groupoid via the differential-geometric fat point a priori only yielads a presheaf of graded manifolds. In this article we prove that this presheaf is representable by the tangent complex of the Lie $n$-groupoid. As an immediate consequence we obtain that the tangent complex carries the structure of a Lie $n$-algebroid.

math.DG↗

arXiv2023

Interface-tuning of ferroelectricity and quadruple-well state in CuInP$_2$S$_6$ via ferroelectric oxide

Ferroelectric van der Waals CuInP$_2$S$_6$ possesses intriguing quadruple-well states and negative piezoelectricity. Its technological implementation has been impeded by the relatively low Curie temperature (bulk $T_C$ ~42 {\deg}C) and the lack of precise domain control. Here we show that CuInP$_2$S$_6$ can be immune to the finite size effect and exhibits enhanced ferroelectricity, piezoelectricity, and polar alignment in the ultrathin limit when interfaced with ferroelectric oxide PbZr$_{0.2}$Ti$_{0.8}$O$_3$ films. Piezoresponse force microscopy studies reveal that the polar domains in thin CuInP$_2$S$_6$ fully conform to those of underlying PbZr$_{0.2}$Ti$_{0.8}$O$_3$, where the piezoelectric coefficient changes sign and increases sharply with reducing thickness. High temperature $in$ $situ$ domain imaging points to a significantly enhanced $T_C$ exceeding 200 {\deg}C for 13 nm CuInP$_2$S$_6$ on PbZr$_{0.2}$Ti$_{0.8}$O$_3$. Density functional theory modeling and Monte Carlo simulations show that the enhanced polar alignment and $T_C$ can be attributed to interface-mediated structure distortion in CuInP$_2$S$_6$. Our study provides an effective material strategy to engineer the polar properties of CuInP$_2$S$_6$ for flexible nanoelectronic, optoelectronic, and mechanical applications.

cond-mat.mtrl-sci↗

arXiv2022

DeepSpeed Inference: Enabling Efficient Inference of Transformer Models at Unprecedented Scale

The past several years have witnessed the success of transformer-based models, and their scale and application scenarios continue to grow aggressively. The current landscape of transformer models is increasingly diverse: the model size varies drastically with the largest being of hundred-billion parameters; the model characteristics differ due to the sparsity introduced by the Mixture-of-Experts; the target application scenarios can be latency-critical or throughput-oriented; the deployment hardware could be single- or multi-GPU systems with different types of memory and storage, etc. With such increasing diversity and the fast-evolving pace of transformer models, designing a highly performant and efficient inference system is extremely challenging. In this paper, we present DeepSpeed Inference, a comprehensive system solution for transformer model inference to address the above-mentioned challenges. DeepSpeed Inference consists of (1) a multi-GPU inference solution to minimize latency while maximizing the throughput of both dense and sparse transformer models when they fit in aggregate GPU memory, and (2) a heterogeneous inference solution that leverages CPU and NVMe memory in addition to the GPU memory and compute to enable high inference throughput with large models which do not fit in aggregate GPU memory. DeepSpeed Inference reduces latency by up to 7.3X over the state-of-the-art for latency-oriented scenarios and increases throughput by over 1.5x for throughput-oriented scenarios. Moreover, it enables trillion parameter scale inference under real-time latency constraints by leveraging hundreds of GPUs, an unprecedented scale for inference. It can inference 25x larger models than with GPU-only solutions, while delivering a high throughput of 84 TFLOPS (over $50\%$ of A6000 peak).

cs.LG↗

arXiv2015

Higher Groupoid Actions, Bibundles, and Differentiation

In this thesis, we employ simplicial methods to study actions, principal bundles, and bibundles of higher groupoids. Roughly, we use Kan fibrations to model actions of higher groupoids, we use pairs of a Kan fibration and a special acyclic fibration to model principal bundles of higher groupoids, we use inner Kan fibrations over the interval to model bibundles of higher groupoids. In particular, we show that our definitions given by the simplicial method agree with those given by the categorification approach to actions, principal bundles, and bibundles of 2-groupoids. In addition, we use the simplicial technique to prove a theorem on differentiation of higher Lie groupoids, which shows that the differentiation functor sends a higher Lie groupoid to a higher Lie algebroid.

math.DG↗