arXiv subjects

Ming Gong

Publications and source records attributed to Ming Gong.

At least 37 records · Page 2Linked to original sources

arXiv2025

Algorithmic Quantum Simulations of Quantum Thermodynamics

Characterizing quantum phases-of-matter at finite-temperature is essential for understanding complex materials and large-scale thermodynamic phenomena. Here, we develop algorithmic protocols for simulating quantum thermodynamics on quantum hardware through quantum kernel function expansion (QKFE), producing the free energy as an analytic function of temperature with uniform convergence. These protocols are demonstrated by simulating transverse field Ising and XY models with superconducting qubits. In both analogue and digital implementations of the QKFE algorithms, we exhibit quantitative agreement of our quantum simulation experiments with the exact results. Our approach provides a general framework for computing thermodynamic potentials on programmable quantum devices, granting access to key thermodynamic properties such as entropy, heat capacity and criticality, with far-reaching implications for material design and drug development.

quant-ph↗

arXiv2025

Coupling between vibration and Luttinger liquid in mechanical nanowires

The vibration of the mechanical nanowire coupled to photons via photon pressure and coupled to charges via the capacity has been widely explored in experiments in the past decades. This system is electrically neutral, thus its coupling to the other degrees of freedom is always challenging. Here, we show that the vibration can slightly change the nanowire length and the associated Fermi velocity, which leads to coupling between vibration and Luttinger liquid. We consider the transverse and longitudinal vibrations of the nanowires, showing that the transverse vibration is much more significant than the longitudinal vibration, which can be measured through the sizable frequency shift. We predict an instability of the vibration induced by this coupling when the frequency becomes negative at a critical temperature for the transverse vibrations in nanowires with low Fermi energy, which can be reached by tuning the chemical potential and magnetic field. The time-dependent oscillation of the conductance, which directly measures the Luttinger parameter, can provide evidence for this coupling. Our theory offers a new mechanism for exploring the coupling between the vibration and the electronic excitations, which may lead to intriguing applications in cooling and controlling the mechanical oscillators with currents.

cond-mat.mes-hall↗

arXiv2025

ECom-Bench: Can LLM Agent Resolve Real-World E-commerce Customer Support Issues?

In this paper, we introduce ECom-Bench, the first benchmark framework for evaluating LLM agent with multimodal capabilities in the e-commerce customer support domain. ECom-Bench features dynamic user simulation based on persona information collected from real e-commerce customer interactions and a realistic task dataset derived from authentic e-commerce dialogues. These tasks, covering a wide range of business scenarios, are designed to reflect real-world complexities, making ECom-Bench highly challenging. For instance, even advanced models like GPT-4o achieve only a 10-20% pass^3 metric in our benchmark, highlighting the substantial difficulties posed by complex e-commerce scenarios. The code and data have been made publicly available at https://github.com/XiaoduoAILab/ECom-Bench to facilitate further research and development in this domain.

cs.CL↗

arXiv2025

Skin mode tunability and self-healing effect in photonic Floquet lattices

Non-Hermitian systems host exotic phenomena absent in their Hermitian counterparts, including the recently predicted self-healing effect (SHE) of non-Hermitian skin modes. To date, the SHE of skin modes in non-Hermitian systems has not been observed experimentally. Here we propose a feasible scheme to realize SHE in photonic Floquet lattices by exploiting skin mode tunability (SMT), a mechanism in which the spectrum of skin modes localized at one boundary can be tuned via a potential applied at the opposite boundary. Such tunability arises from the non-Hermitian biorthogonality of the eigenstates. We demonstrate that a certain skin mode is exceptionally sensitive to remote-boundary potentials in an array of $100$ coupled helical waveguides, allowing broad-range spectral control and the generation of SHE with experimentally accessible parameters. Our results establish a general framework for engineering skin modes via local perturbations, thereby expanding the toolbox for non-Hermitian wave control.

physics.optics↗

arXiv2025

Inverse designed Hamiltonians for perfect state transfer and remote entanglement generation, and applications in superconducting qubits

Hamiltonian inverse engineering enables the design of protocols for specific quantum evolutions or target state preparation. Perfect state transfer (PST) and remote entanglement generation are notable examples, as they serve as key primitives in quantum information processing. However, Hamiltonians obtained through conventional methods often lack robustness against noise. Assisted by inverse engineering, we begin with a noise-resilient energy spectrum and construct a class of Hamiltonians, referred to as the dome model, that significantly improves the system's robustness against noise, as confirmed by numerical simulations. This model introduces a tunable parameter $m$ that modifies the energy-level spacing and gives rise to a well-structured Hamiltonian. It reduces to the conventional PST model at $m=0$ and simplifies to a SWAP model involving only two end qubits in the large-$m$ regime. To address the challenge of scalability, we propose a cascaded strategy that divides long-distance PST into multiple consecutive PST steps. Our work is particularly suited for demonstration on superconducting qubits with tunable couplers, which enable rapid and flexible Hamiltonian engineering, thereby advancing the experimental potential of robust and scalable quantum information processing.

quant-ph↗

arXiv2025

Proof of the exact diffusion constant via first passage time in quasi-periodic potentials

Brownian motion in terms of Lifson and Jackson (LJ) formula has been widely explored in periodic systems and it has been believed for a long time that the LJ formula only applies to periodic potentials. Recently we show that for the following Brownian motion $γ\dot{x} = -U'(x) + ξ$, where $U(x)$ is the quasi-periodic potential, the effective diffusion constant can still be described by the LJ formula $D^* = D/(\langle \exp(βU)\rangle \langle \exp(-βU)\rangle)$, where the average is redefined as $\langle \exp(βU)\rangle = \lim_{L\rightarrow \infty} L^{-1} \int_0^L \exp(βU(x))dx$. In this manuscript we prove this result exactly using the mean first passage time $τ(x)$, with boundary conditions $τ(\pm L) = 0$, and show that the effective diffusion constant can be determined using $D^* =\lim_{L \rightarrow \infty} L^2/(2τ(0))$, where $\pm L$ is the two positions of the absorbing boundary. We exactly solve the equation of motion of $τ(x)$ and obtain the above result with the aid of Jacobi-Anger expansion method. Our result can be generalized to the other potentials and even higher dimensions, which can greatly broaden our understanding of Brownian motion in more general circumstances. The requirement for a well-defined effective diffusion constant $D^*$ in more general potentials is also discussed.

cond-mat.stat-mech↗

arXiv2025

Manipulating topological charges via engineering zeros of wave functions

Topological charges are typically manipulated by managing their energy bands in quantum systems. In this work, we propose a new approach to manipulate the topological charges of systems by engineering density zeros of localized wave excitations in them. We demonstrate via numerical simulation and analytical analysis that the winding number of a toroidal Bose condensate can be well manipulated by engineering the relative velocities between the dark solitons and their backgrounds. The crossing of relative velocities through zero makes a change in winding number by inducing density zeros during acceleration, with the direction of crossing determining whether charge increases or decreases. Possibilities of observing such winding number manipulation are discussed for current experimental settings. This idea may also be to higher dimensions. These results will inspire new pathways in designing topological materials using quantum simulation platforms.

cond-mat.quant-gas↗

arXiv2025

Spontaneous formation of subsystem and bath under accordion-type driving

Floquet modulations often yield effective Hamiltonians not easily accessible in traditional time-dependent systems, which brings opportunities for exploring novel physics of quantum dynamics. We investigate a Floquet system exhibiting translational symmetry at any fixed time but the spatial periodicity is time-dependent. Such a system is a natural platform for studying thermalization and novel dynamical structures. We find that the single-particle Hilbert space spontaneously develops a structure of a two-level subsystem and the rest part forms a bath. The dynamic process is analyzed perturbatively within the two-level subsystem as well as numerical solutions, exhibiting stable time-evolutions. These results enrich our understanding of Floquet thermalization without definite spatial periodicity, which brings hints for exploring many-body physics such as scar states.

quant-ph↗

arXiv2025

Modeling Sequential Sentence Relation to Improve Cross-lingual Dense Retrieval

Recently multi-lingual pre-trained language models (PLM) such as mBERT and XLM-R have achieved impressive strides in cross-lingual dense retrieval. Despite its successes, they are general-purpose PLM while the multilingual PLM tailored for cross-lingual retrieval is still unexplored. Motivated by an observation that the sentences in parallel documents are approximately in the same order, which is universal across languages, we propose to model this sequential sentence relation to facilitate cross-lingual representation learning. Specifically, we propose a multilingual PLM called masked sentence model (MSM), which consists of a sentence encoder to generate the sentence representations, and a document encoder applied to a sequence of sentence vectors from a document. The document encoder is shared for all languages to model the universal sequential sentence relation across languages. To train the model, we propose a masked sentence prediction task, which masks and predicts the sentence vector via a hierarchical contrastive loss with sampled negatives. Comprehensive experiments on four cross-lingual retrieval tasks show MSM significantly outperforms existing advanced pre-training models, demonstrating the effectiveness and stronger cross-lingual retrieval capabilities of our approach. Code and model are available at https://github.com/shunyuzh/MSM.

cs.CL↗

arXiv2025

Unilateral Criticality and Phase Transition in the Cavity-Ising Model

Superradiant phase transitions from cavity light-matter coupling have been widely explored across platforms. Here, we report a unilateral critical endpoint (UCEP) and a tricritical point (TCP) in the phase diagram of the cavity-coupled transverse Ising model with $\mathbb{Z}_2$ symmetry. At zero temperature, we demonstrate that this model hosts three phases separated by two second-order and one first-order transitions. These lines intersect at a TCP and a UCEP, the latter not captured by existing phase-transition paradigms. The UCEP displays one-sided criticality: approaching the point from one side, the system behaves as a second-order transition, while from the other side it is first-order. Correspondingly, two order parameters, respectively, undergo the first- and the second-order phase transitions at the same point. We construct a minimal description of UCEP with the density of the free energy $f = c_{1}(\tildeα^{2}+c_{2})+(\tildeα^{2}+c_{2})^{2}\ln{\vert\tildeα^{2}+c_{2}\vert}$, with the UCEP at $(c_{1},c_{2})=(1/e,0)$ and $\tildeα$ being the order parameter. We further map the finite-temperature phase diagram and perform a symmetry analysis. By unifying first- and second-order signatures in a single, direction-dependent endpoint, the UCEP introduces a qualitatively new class of phase transition and may have applications in fields such as quantum measurement and quantum sensing. This work also provides an intriguing platform for exploring novel critical phenomena in cavity-coupled many-body systems with or without dissipation.

quant-ph↗

arXiv2025

Structure-Learnable Adapter Fine-Tuning for Parameter-Efficient Large Language Models

This paper addresses the issues of parameter redundancy, rigid structure, and limited task adaptability in the fine-tuning of large language models. It proposes an adapter-based fine-tuning method built on a structure-learnable mechanism. By introducing differentiable gating functions and structural sparsity control variables, the method enables automatic optimization of adapter insertion points, activation paths, and module combinations. This allows the model to adjust its structure flexibly in multi-task settings to match different task characteristics. With the backbone parameters kept frozen, the method uses a structure search mechanism to guide the dynamic construction of task-specific efficient substructures during training. This significantly improves parameter utilization and representational capacity. In addition, the paper designs a set of sensitivity analysis experiments to systematically evaluate the effects of sparsity weight, noise injection ratio, and data perturbation on model performance. These experiments verify the stability and robustness of the proposed method across various multi-task natural language understanding tasks. The experimental results show that the proposed method outperforms mainstream parameter-efficient tuning techniques on multiple tasks. It achieves a better balance among accuracy, compression rate, and robustness to noise and perturbation.

cs.CL↗

arXiv2025

Graph Neural Network and Transformer Integration for Unsupervised System Anomaly Discovery

This study proposes an unsupervised anomaly detection method for distributed backend service systems, addressing practical challenges such as complex structural dependencies, diverse behavioral evolution, and the absence of labeled data. The method constructs a dynamic graph based on service invocation relationships and applies graph convolution to extract high-order structural representations from multi-hop topologies. A Transformer is used to model the temporal behavior of each node, capturing long-term dependencies and local fluctuations. During the feature fusion stage, a learnable joint embedding mechanism integrates structural and behavioral representations into a unified anomaly vector. A nonlinear mapping is then applied to compute anomaly scores, enabling an end-to-end detection process without supervision. Experiments on real-world cloud monitoring data include sensitivity analyses across different graph depths, sequence lengths, and data perturbations. Results show that the proposed method outperforms existing models on several key metrics, demonstrating stronger expressiveness and stability in capturing anomaly propagation paths and modeling dynamic behavior sequences, with high potential for practical deployment.

cs.LG↗

arXiv2025

Multi-Level Service Performance Forecasting via Spatiotemporal Graph Neural Networks

This paper proposes a spatiotemporal graph neural network-based performance prediction algorithm to address the challenge of forecasting performance fluctuations in distributed backend systems with multi-level service call structures. The method abstracts system states at different time slices into a sequence of graph structures. It integrates the runtime features of service nodes with the invocation relationships among services to construct a unified spatiotemporal modeling framework. The model first applies a graph convolutional network to extract high-order dependency information from the service topology. Then it uses a gated recurrent network to capture the dynamic evolution of performance metrics over time. A time encoding mechanism is also introduced to enhance the model's ability to represent non-stationary temporal sequences. The architecture is trained in an end-to-end manner, optimizing the multi-layer nested structure to achieve high-precision regression of future service performance metrics. To validate the effectiveness of the proposed method, a large-scale public cluster dataset is used. A series of multi-dimensional experiments are designed, including variations in time windows and concurrent load levels. These experiments comprehensively evaluate the model's predictive performance and stability. The experimental results show that the proposed model outperforms existing representative methods across key metrics such as MAE, RMSE, and R2. It maintains strong robustness under varying load intensities and structural complexities. These results demonstrate the model's practical potential for backend service performance management tasks.

cs.LG↗

arXiv2025

MindFlow+: A Self-Evolving Agent for E-Commerce Customer Service

High-quality dialogue is crucial for e-commerce customer service, yet traditional intent-based systems struggle with dynamic, multi-turn interactions. We present MindFlow+, a self-evolving dialogue agent that learns domain-specific behavior by combining large language models (LLMs) with imitation learning and offline reinforcement learning (RL). MindFlow+ introduces two data-centric mechanisms to guide learning: tool-augmented demonstration construction, which exposes the model to knowledge-enhanced and agentic (ReAct-style) interactions for effective tool use; and reward-conditioned data modeling, which aligns responses with task-specific goals using reward signals. To evaluate the model's role in response generation, we introduce the AI Contribution Ratio, a novel metric quantifying AI involvement in dialogue. Experiments on real-world e-commerce conversations show that MindFlow+ outperforms strong baselines in contextual relevance, flexibility, and task accuracy. These results demonstrate the potential of combining LLMs tool reasoning, and reward-guided learning to build domain-specialized, context-aware dialogue systems.

cs.CL↗

arXiv2025

Autonomous Resource Management in Microservice Systems via Reinforcement Learning

This paper proposes a reinforcement learning-based method for microservice resource scheduling and optimization, aiming to address issues such as uneven resource allocation, high latency, and insufficient throughput in traditional microservice architectures. In microservice systems, as the number of services and the load increase, efficiently scheduling and allocating resources such as computing power, memory, and storage becomes a critical research challenge. To address this, the paper employs an intelligent scheduling algorithm based on reinforcement learning. Through the interaction between the agent and the environment, the resource allocation strategy is continuously optimized. In the experiments, the paper considers different resource conditions and load scenarios, evaluating the proposed method across multiple dimensions, including response time, throughput, resource utilization, and cost efficiency. The experimental results show that the reinforcement learning-based scheduling method significantly improves system response speed and throughput under low load and high concurrency conditions, while also optimizing resource utilization and reducing energy consumption. Under multi-dimensional resource conditions, the proposed method can consider multiple objectives and achieve optimized resource scheduling. Compared to traditional static resource allocation methods, the reinforcement learning model demonstrates stronger adaptability and optimization capability. It can adjust resource allocation strategies in real time, thereby maintaining good system performance in dynamically changing load and resource environments.

cs.DC↗

arXiv2025

Soliton and traveling wave solutions in coupled one-dimensional condensates

Ultracold condensates provide a unique platform for exploring soliton physics. Motivated by the recent experiments realizing the sine-Gordon model in a split one-dimensional (1D) BEC, we demonstrate that this system naturally supports various density and phase solitons. We explore the physics using the bosonization technique, in which the phase and density are conjugate pairs, and determine its effective Language equation and the associated equation of motion. We show that in the presence of asymmetry between the two condensates, new solutions beyond those in the sine-Gordon model emerge. We calculate the traveling wave solutions and soliton solutions in this model and determine their corresponding energy densities analytically. Finally, we discuss the relevance of these solutions to the experiments and discuss their observations. This theory does not rely on the mechanism of quasi-particle excitation, which yields the Lee-Huang-Yang correction in higher dimensions, and is thus much more suitable to describe the physics in 1D systems. Since the physical models have already been realized in experiments, this work opens a new frontier for the realization of various soliton and periodic solutions using two coupled condensates.

cond-mat.quant-gas↗

arXiv2025

MindFlow: Revolutionizing E-commerce Customer Support with Multimodal LLM Agents

Recent advances in large language models (LLMs) have enabled new applications in e-commerce customer service. However, their capabilities remain constrained in complex, multimodal scenarios. We present MindFlow, the first open-source multimodal LLM agent tailored for e-commerce. Built on the CoALA framework, it integrates memory, decision-making, and action modules, and adopts a modular "MLLM-as-Tool" strategy for effect visual-textual reasoning. Evaluated via online A/B testing and simulation-based ablation, MindFlow demonstrates substantial gains in handling complex queries, improving user satisfaction, and reducing operational costs, with a 93.53% relative improvement observed in real-world deployments.

cs.CL↗

arXiv2025

Emergence of Nodal-Knot Transitions by Disorder

Under certain symmetries, degenerate points in three-dimensional metals form one-dimensional nodal lines. These nodal lines sometimes exhibit intricate knotted structures and have been studied in various contexts. As one of the most common physical perturbations, disorder effects often trigger novel quantum phase transitions. For nodal-knot phases, whether disorder can drive knot transitions remains an open and intriguing question. Employing renormalization-group calculations, we demonstrate that nodal-knot transitions emerge in the presence of weak disorder. Specifically, both chemical-potential-type and magnetic-type disorders can induce knot transitions, resulting in the emergence of distinct knot topologies. The transition can be quantitatively characterized by changes in topological invariants such as the knot Wilson loop integrals. Our findings open up a new avenue for manipulating the topology of nodal-knot phases through disorder effects.

cond-mat.mes-hall↗