Searcharxiv⌕ Search

arXiv subjects

Qi Zhou

Publications and source records attributed to Qi Zhou.

At least 91 records · Page 5Linked to original sources
arXiv2025

Polarization-dependent chiral transport and chiral solitons in spin Kitaev models

Recent advances in synthetic quantum matter allow researchers to design quantum models inaccessible in traditional materials. Here, we propose protocols to engineer a new class of quantum spin models, which we call spin Kitaev models. The building blocks are basic spin-exchange interactions combined with locally selective Floquet pulses, a capability recently demonstrated in a range of experimental platforms. The resulting flip-flip and flop-flop terms lead to intriguing quantum transport dynamics beyond conventional spin models. For instance, in the absence of a magnetic field, spin excitations polarized along the $x$ and $y$ axes propagate chirally in opposite directions, producing polarization-dependent spin transport. In the large-spin limit, the spin Kitaev model maps to a nonlinear Hatano-Nelson model, where the interplay of nonlinearity and the underlying curvature yields polarization-dependent chiral solitons. A magnetic field binds two oppositely polarized chiral solitons into a chiral solitonic molecule, whose travel direction depends on its orientation. Our results, directly accessible in current experiments, open new opportunities for simulating transport in curved spaces and for applications in spintronics, information processing, and quantum sensing.

cond-mat.quant-gas↗
arXiv2025

Low Power, Scalable Nanofabrication via Photon Upconversion

Micro- and nanoscale fabrication, which enables precise construction of intricate three-dimensional structures, is of foundational importance for advancing innovation in plasmonics, nanophotonics, and biomedical applications. However, scaling fabrication to industrially relevant levels remains a significant challenge. We demonstrate that triplet-triplet annihilation upconversion (TTA-UC) offers a unique opportunity to increase fabrication speeds and scalability of micro- and nanoscale 3D structures. Due to its nonlinearity and low power requirements, TTA-UC enables localized polymerization with nanoscale resolutions while simultaneously printing millions of voxels per second through optical parallelization using off-the-shelf light-emitting diodes and digital micromirror devices. Our system design and component integration empower fabrication with a minimum lateral feature size down to 230 nm and speeds up to 112 million voxels per second at a power of 7.0 nW per voxel. This combination of high resolution and fast print speed demonstrates that TTA-UC is a significant advancement in nanofabrication technique, evidenced by the fabrication of hydrophobic nanostructures on a square-centimeter scale, paving the way for industrial nanomanufacturing.

physics.app-ph↗
arXiv2025

Large Foundation Model for Ads Recommendation

Online advertising relies on accurate recommendation models, with recent advances using pre-trained large-scale foundation models (LFMs) to capture users' general interests across multiple scenarios and tasks. However, existing methods have critical limitations: they extract and transfer only user representations (URs), ignoring valuable item representations (IRs) and user-item cross representations (CRs); and they simply use a UR as a feature in downstream applications, which fails to bridge upstream-downstream gaps and overlooks more transfer granularities. In this paper, we propose LFM4Ads, an All-Representation Multi-Granularity transfer framework for ads recommendation. It first comprehensively transfers URs, IRs, and CRs, i.e., all available representations in the pre-trained foundation model. To effectively utilize the CRs, it identifies the optimal extraction layer and aggregates them into transferable coarse-grained forms. Furthermore, we enhance the transferability via multi-granularity mechanisms: non-linear adapters for feature-level transfer, an Isomorphic Interaction Module for module-level transfer, and Standalone Retrieval for model-level transfer. LFM4Ads has been successfully deployed in Tencent's industrial-scale advertising platform, processing tens of billions of daily samples while maintaining terabyte-scale model parameters with billions of sparse embedding keys across approximately two thousand features. Since its production deployment in Q4 2024, LFM4Ads has achieved 10+ successful production launches across various advertising scenarios, including primary ones like Weixin Moments and Channels. These launches achieve an overall GMV lift of 2.45% across the entire platform, translating to estimated annual revenue increases in the hundreds of millions of dollars.

cs.LG↗
arXiv2025

InfiAlign: A Scalable and Sample-Efficient Framework for Aligning LLMs to Enhance Reasoning Capabilities

Large language models (LLMs) have exhibited impressive reasoning abilities on a wide range of complex tasks. However, enhancing these capabilities through post-training remains resource intensive, particularly in terms of data and computational cost. Although recent efforts have sought to improve sample efficiency through selective data curation, existing methods often rely on heuristic or task-specific strategies that hinder scalability. In this work, we introduce InfiAlign, a scalable and sample-efficient post-training framework that integrates supervised fine-tuning (SFT) with Direct Preference Optimization (DPO) to align LLMs for enhanced reasoning. At the core of InfiAlign is a robust data selection pipeline that automatically curates high-quality alignment data from open-source reasoning datasets using multidimensional quality metrics. This pipeline enables significant performance gains while drastically reducing data requirements and remains extensible to new data sources. When applied to the Qwen2.5-Math-7B-Base model, our SFT model achieves performance on par with DeepSeek-R1-Distill-Qwen-7B, while using only approximately 12% of the training data, and demonstrates strong generalization across diverse reasoning tasks. Additional improvements are obtained through the application of DPO, with particularly notable gains in mathematical reasoning tasks. The model achieves an average improvement of 3.89% on AIME 24/25 benchmarks. Our results highlight the effectiveness of combining principled data selection with full-stage post-training, offering a practical solution for aligning large reasoning models in a scalable and data-efficient manner. The model checkpoints are available at https://huggingface.co/InfiX-ai/InfiAlign-Qwen-7B-SFT.

cs.AI↗
arXiv2025

Observation of Metal-Insulator and Spectral Phase Transitions in Aubry-André-Harper Models

Non-Hermitian extensions of the Aubry-André-Harper (AAH) model reveal a rich variety of phase transitions arising from the interplay of quasiperiodicity and non-Hermiticity. Despite their theoretical significance, experimental explorations remain challenging due to complexities in realizing controlled non-Hermiticity. Here, we present the first experimental realization of the unitary almost-Mathieu operator (UAMO) which simulates the AAH model by employing single-photon quantum walks. Through precise control of quasiperiodicity, we systematically explore the phase diagram displaying a phase transition between localized and delocalized regimes in the Hermitian limit. Subsequently, by introducing non-reciprocal hopping, we experimentally probe the parity-time (PT) symmetry-breaking transition that is characterized by the emergence of complex quasienergies. Moreover, we identify a novel spectral transition exclusive to discrete-time settings, where all quasienergies become purely imaginary. Both transitions are connected to changes in the spectral winding number, demonstrating their topological origins. These results clarify the interplay between localization, symmetry breaking, and topology in non-Hermitian quasicrystals, paving the way for future exploration of synthetic quantum matter.

quant-ph↗
arXiv2025

An RRT* algorithm based on Riemannian metric model for optimal path planning

This paper presents a Riemannian metric-based model to solve the optimal path planning problem on two-dimensional smooth submanifolds in high-dimensional space. Our model is based on constructing a new Riemannian metric on a two-dimensional projection plane, which is induced by the high-dimensional Euclidean metric on two-dimensional smooth submanifold and reflects the environmental information of the robot. The optimal path planning problem in high-dimensional space is therefore transformed into a geometric problem on the two-dimensional plane with new Riemannian metric. Based on the new Riemannian metric, we proposed an incremental algorithm RRT*-R on the projection plane. The experimental results show that the proposed algorithm is suitable for scenarios with uneven fields in multiple dimensions. The proposed algorithm can help the robot to effectively avoid areas with drastic changes in height, ground resistance and other environmental factors. More importantly, the RRT*-R algorithm shows better smoothness and optimization properties compared with the original RRT* algorithm using Euclidean distance in high-dimensional workspace. The length of the entire path by RRT*-R is a good approximation of the theoretical minimum geodesic distance on projection plane.

cs.RO↗
arXiv2025

Subordinacy theory for long-range operators: hyperbolic geodesic flow insights and monotonicity theory

We introduce a comprehensive framework for subordinacy theory applicable to long-range operators on $\ell^2(\mathbb Z)$, bridging dynamical systems and spectral analysis. For finite-range operators, we establish a correspondence between the dynamical behavior of partially hyperbolic (Hermitian-)symplectic cocycles and the existence of purely absolutely continuous spectrum, resolving an open problem posed by Jitomirskaya. For infinite-range operators-where traditional cocycle methods become inapplicable-we characterize absolutely continuous spectrum through the growth of generalized eigenfunctions, extending techniques from higher-dimensional lattice models. Our main results include the first rigorous proof of purely absolutely continuous spectrum for quasi-periodic long-range operators with analytic potentials and Diophantine frequencies-in particular, the first proof of the all-phases persistence for finite-range perturbations of subcritical almost Mathieu operators-among other advances in spectral theory of long-range operators. The key novelty of our approach lies in the unanticipated connection between stable/vertical bundle intersections in geodesic flows-where they detect conjugate points-and their equally fundamental role in governing (de-)localization for Schrödinger operators. The geometric insight, combined with a novel coordinate-free monotonicity theory for general bundles (including its preservation under center-bundle restrictions) and adapted analytic spectral and KAM techniques, enables our spectral analysis of long-range operators.

math.DS↗
arXiv2025

Cantor spectrum for multidimensional quasi-periodic Schrödinger operators

In this paper, we investigate the spectrum of a class of multidimensional quasi-periodic Schrödinger operators that exhibit a Cantor spectrum, which provides a resolution to a question posed by Damanik, Fillman, and Gorodetski \cite{DFG}. Additionally, we prove that for a dense set of irrational frequencies with positive Hausdorff dimension, the Hausdorff (and upper box) dimension of the spectrum of the critical almost Mathieu operator is positive, yet can be made arbitrarily small.

math.SP↗
arXiv2025

Electromagnetic emission from strongly interacting hadronic and partonic matter created in heavy-ion collisions

We investigate dilepton production in heavy-ion, proton-proton, and proton-nucleus collisions from low energies of 1 AGeV (SIS) to ultra-relativistic energies (LHC) using the Parton-Hadron-String Dynamics (PHSD) transport approach. PHSD is a microscopic, non-equilibrium approach that integrates hadronic and partonic degrees of freedom, providing a comprehensive description of relativistic heavy-ion collisions from initial nucleon-nucleon interactions to quark-gluon plasma (QGP) formation, hadronization, and final-state interactions. Key dilepton sources in PHSD include hadronic decays, bremsstrahlung, QGP radiation ($q+\bar q \to e^+e^-$, \ $q+\bar q \to g+ e^+e^-$, \ $q+g \to q+ e^+e^-$), primary Drell-Yan production, and semileptonic decays of correlated charm and bottom pairs. PHSD well describes dilepton data from HADES, STAR, and ALICE experiments. We examine in-medium effects, such as the vector meson spectral function broadening, and present the excitation function for the dilepton "excess" in the invariant mass range $0.4<M_{ee}<0.75$ GeV/c$^2$. For the first time we report on the baryon chemical potential $μ_B$-dependence of the QGP radiation calculated on a basis of the Dynamical-Quasi-Particle Model (DQPM) in PHSD. The influence of $μ_B$ on the QGP yield grows at lower collision energies, where $μ_B$ becomes large, however, its impact on the total dilepton spectra is small due to the lowering of the QGP volume with decreasing energy. The excitation function of QGP dileptons, is presented versus correlated charm, confirming that the QGP radiation overshines charm contributions at $\sqrt{s} \simeq 25-30$ GeV in central Au+Au collisions, providing access to thermal QGP dileptons at BES RHIC and FAIR.

nucl-th↗
arXiv2025

Fair-PP: A Synthetic Dataset for Aligning LLM with Personalized Preferences of Social Equity

Human preference plays a crucial role in the refinement of large language models (LLMs). However, collecting human preference feedback is costly and most existing datasets neglect the correlation between personalization and preferences. To address this issue, we introduce Fair-PP, a synthetic dataset of personalized preferences targeting social equity, derived from real-world social survey data, which includes 28 social groups, 98 equity topics, and 5 personal preference dimensions. Leveraging GPT-4o-mini, we engage in role-playing based on seven representative persona portrayals guided by existing social survey data, yielding a total of 238,623 preference records. Through Fair-PP, we also contribute (i) An automated framework for generating preference data, along with a more fine-grained dataset of personalized preferences; (ii) analysis of the positioning of the existing mainstream LLMs across five major global regions within the personalized preference space; and (iii) a sample reweighting method for personalized preference alignment, enabling alignment with a target persona while maximizing the divergence from other personas. Empirical experiments show our method outperforms the baselines.

cs.AI↗
arXiv2025

DFVO: Learning Darkness-free Visible and Infrared Image Disentanglement and Fusion All at Once

Visible and infrared image fusion is one of the most crucial tasks in the field of image fusion, aiming to generate fused images with clear structural information and high-quality texture features for high-level vision tasks. However, when faced with severe illumination degradation in visible images, the fusion results of existing image fusion methods often exhibit blurry and dim visual effects, posing major challenges for autonomous driving. To this end, a Darkness-Free network is proposed to handle Visible and infrared image disentanglement and fusion all at Once (DFVO), which employs a cascaded multi-task approach to replace the traditional two-stage cascaded training (enhancement and fusion), addressing the issue of information entropy loss caused by hierarchical data transmission. Specifically, we construct a latent-common feature extractor (LCFE) to obtain latent features for the cascaded tasks strategy. Firstly, a details-extraction module (DEM) is devised to acquire high-frequency semantic information. Secondly, we design a hyper cross-attention module (HCAM) to extract low-frequency information and preserve texture features from source images. Finally, a relevant loss function is designed to guide the holistic network learning, thereby achieving better image fusion. Extensive experiments demonstrate that our proposed approach outperforms state-of-the-art alternatives in terms of qualitative and quantitative evaluations. Particularly, DFVO can generate clearer, more informative, and more evenly illuminated fusion results in the dark environments, achieving best performance on the LLVIP dataset with 63.258 dB PSNR and 0.724 CC, providing more effective information for high-level vision tasks. Our code is publicly accessible at https://github.com/DaVin-Qi530/DFVO.

cs.CV↗
arXiv2025

Weighted balanced truncation method for approximating kernel functions by exponentials

Kernel approximation with exponentials is useful in many problems with convolution quadrature and particle interactions such as integral-differential equations, molecular dynamics and machine learning. This paper proposes a weighted balanced truncation to construct an optimal model reduction method for compressing the number of exponentials in the sum-of-exponentials approximation of kernel functions. This method shows great promise in approximating long-range kernels, achieving over 4 digits of accuracy improvement for the Ewald-splitting and inverse power kernels in comparison with the classical balanced truncation. Numerical results demonstrate its excellent performance and attractive features for practical applications.

physics.comp-ph↗
arXiv2025

Nonlocal diffusion and pulse intervention in a faecal-oral model with moving infected fronts

How individual dispersal patterns and human intervention behaviours affect the spread of infectious diseases constitutes a central problem in epidemiological research. This paper develops an impulsive nonlocal faecal-oral model with free boundaries, where pulses are introduced to capture a periodic spraying of disinfectant, and nonlocal diffusion describes the long-range dispersal of individuals, and free boundaries represent moving infected fronts. We first check that the model has a unique nonnegative global classical solution. Then, the principal eigenvalue, which depends on the infected region, the impulse intensity, and the kernel functions for nonlocal diffusion, is examined by using the theory of resolvent positive operators and their perturbations. Based on this value, this paper obtains that the diseases are either vanishing or spreading, and provides criteria for determining when vanishing and spreading occur. At the end, a numerical example is presented in order to corroborate the theoretical findings and to gain further understanding of the effect of the pulse intervention. This work shows that the pulsed intervention is beneficial in combating the diseases, but the effect of the nonlocal diffusion depends on the choice of the kernel functions.

math.AP↗
arXiv2025

Accurate Error Estimates and Optimal Parameter Selection in Ewald Summation for Dielectrically Confined Coulomb Systems

Dielectrically confined Coulomb systems are widely employed in molecular dynamics (MD) simulations. Despite extensive efforts in developing efficient and accurate algorithms for these systems, rigorous and accurate error estimates, which are crucial for optimal parameter selection for simulations, is still lacking. In this work, we present a rigorous error analysis in Ewald summation for electrostatic interactions in systems with two dielectric planar interfaces, where the polarization contribution is modeled by an infinitely reflected image charge series. Accurate error estimate is provided for the truncation error of image charge series, as well as decay rates of energy and force correction terms, as functions of system parameters such as vacuum layer thickness, dielectric contrasts, and image truncation levels. Extensive numerical tests conducted across several prototypical parameter settings validate our theoretical predictions. Additionally, our analysis elucidates the non-monotonic error convergence behavior observed in previous numerical studies. Finally, we provide an optimal parameter selection strategy derived from our theoretical insights, offering practical guidance for efficient and accurate MD simulations of dielectric-confined systems.

math.NA↗
arXiv2025

A Novel Solution for Drone Photogrammetry with Low-overlap Aerial Images using Monocular Depth Estimation

Low-overlap aerial imagery poses significant challenges to traditional photogrammetric methods, which rely heavily on high image overlap to produce accurate and complete mapping products. In this study, we propose a novel workflow based on monocular depth estimation to address the limitations of conventional techniques. Our method leverages tie points obtained from aerial triangulation to establish a relationship between monocular depth and metric depth, thus transforming the original depth map into a metric depth map, enabling the generation of dense depth information and the comprehensive reconstruction of the scene. For the experiments, a high-overlap drone dataset containing 296 images is processed using Metashape to generate depth maps and DSMs as ground truth. Subsequently, we create a low-overlap dataset by selecting 20 images for experimental evaluation. Results demonstrate that while the recovered depth maps and resulting DSMs achieve meter-level accuracy, they provide significantly better completeness compared to traditional methods, particularly in regions covered by single images. This study showcases the potential of monocular depth estimation in low-overlap aerial photogrammetry.

cs.CV↗
arXiv2025

Heavy quark mass near the phase transition

Assuming that the number densities of heavy flavor in hadron gas and in QGP are same at $T_c$, we obtain the effective mass of heavy quark at $T_c$ from the comparison with the hadron resonance gas model which well describes particle yield in heavy-ion collisions. We find that charm quark mass at vanishing baryon chemical potential is around 1.8 GeV which is much heavier than QCD bare mass and close to $D$ meson mass. The mass slightly increases with increasing baryon chemical potential and then decreases. On the other hand, anticharm quark mass constantly decreases with increasing baryon chemical potential. Bottom quark mass has a similar pattern. Extending the hadron resonance gas model to a bit higher temperature beyond $T_c$, the effective masses of charm and bottom quarks decrease with increasing temperature.

hep-ph↗
arXiv2025

InfiFusion: A Unified Framework for Enhanced Cross-Model Reasoning via LLM Fusion

We introduce InfiFusion, an efficient training pipeline designed to integrate multiple domain-specialized Large Language Models (LLMs) into a single pivot model, effectively harnessing the strengths of each source model. Traditional fusion methods either merge model parameters directly or rely on knowledge distillation with rigid assumptions, limiting their flexibility and efficiency. InfiFusion overcomes these limitations by enhancing Universal Logit Distillation (ULD) with Top-K selection and Logits Standardization. We propose two fusion strategies: Pairwise Fusion (InfiFusion$_p$), where each source model knowledge is distilled individually into the pivot model followed by merging and Unified Fusion (InfiFusion$_u$), where knowledge from all source models is distilled simultaneously into the pivot model. InfiFusion outperforms the state-of-the-art models, such as Qwen-2.5-14B-Instruct and Phi-4, across 11 widely applied benchmarks covering reasoning, coding, mathematics, and instruction-following tasks. Notably, InfiFusion achieves this superior performance while significantly reduces computational costs, completing full training with only 160 H800 GPU hours compared to the millions typically required for traditional LLM training.

cs.CL↗