arXiv subjects

Xi Zhang

Publications and source records attributed to Xi Zhang.

At least 145 records · Page 8Linked to original sources

arXiv2025

Reasoning Paths as Signals: Augmenting Multi-hop Fact Verification through Structural Reasoning Progression

The growing complexity of factual claims in real-world scenarios presents significant challenges for automated fact verification systems, particularly in accurately aggregating and reasoning over multi-hop evidence. Existing approaches often rely on static or shallow models that fail to capture the evolving structure of reasoning paths, leading to fragmented retrieval and limited interpretability. To address these issues, we propose a Structural Reasoning framework for Multi-hop Fact Verification that explicitly models reasoning paths as structured graphs throughout both evidence retrieval and claim verification stages. Our method comprises two key modules: a structure-enhanced retrieval mechanism that constructs reasoning graphs to guide evidence collection, and a reasoning-path-guided verification module that incrementally builds subgraphs to represent evolving inference trajectories. We further incorporate a structure-aware reasoning mechanism that captures long-range dependencies across multi-hop evidence chains, enabling more precise verification. Extensive experiments on the FEVER and HoVer datasets demonstrate that our approach consistently outperforms strong baselines, highlighting the effectiveness of reasoning-path modeling in enhancing retrieval precision and verification accuracy.

cs.AI↗

arXiv2025

Loki's Dance of Illusions: A Comprehensive Survey of Hallucination in Large Language Models

Edgar Allan Poe noted, "Truth often lurks in the shadow of error," highlighting the deep complexity intrinsic to the interplay between truth and falsehood, notably under conditions of cognitive and informational asymmetry. This dynamic is strikingly evident in large language models (LLMs). Despite their impressive linguistic generation capabilities, LLMs sometimes produce information that appears factually accurate but is, in reality, fabricated, an issue often referred to as 'hallucinations'. The prevalence of these hallucinations can mislead users, affecting their judgments and decisions. In sectors such as finance, law, and healthcare, such misinformation risks causing substantial economic losses, legal disputes, and health risks, with wide-ranging consequences.In our research, we have methodically categorized, analyzed the causes, detection methods, and solutions related to LLM hallucinations. Our efforts have particularly focused on understanding the roots of hallucinations and evaluating the efficacy of current strategies in revealing the underlying logic, thereby paving the way for the development of innovative and potent approaches. By examining why certain measures are effective against hallucinations, our study aims to foster a comprehensive approach to tackling this issue within the domain of LLMs.

cs.CL↗

arXiv2025

Venus Cloud Research: Progress and Perspectives

Venus has regained attention on the international stage with the approval of three new missions by ESA and NASA. As the twin sister of Earth, Venus exhibits a distinct atmosphere, which casts a veil of mystery over the planetary evolution and is of great scientific significance. One of the most important components of Venus-the cloud-is believed to have significantly regulated its climate evolution and affect the environmental habitability. However, due to sparse in-situ measurements and the limitation of remote sensing, properties of these clouds remain largely unknown. Based on research conducted in past decades, this article reviews the observational structure of cloud properties, the progress of microphysical and simplified cloud model developments, and perspectives of future directions of this research field. Several possible solutions to the challenges associated with the coupling effect, ultraviolet absorption, and habitability are proposed and discussed in details, providing insights for future Venus' explorations.

astro-ph.EP↗

arXiv2025

Diffusion with a Linguistic Compass: Steering the Generation of Clinically Plausible Future sMRI Representations for Early MCI Conversion Prediction

Early prediction of Mild Cognitive Impairment (MCI) conversion is hampered by a trade-off between immediacy--making fast predictions from a single baseline sMRI--and accuracy--leveraging longitudinal scans to capture disease progression. We propose MCI-Diff, a diffusion-based framework that synthesizes clinically plausible future sMRI representations directly from baseline data, achieving both real-time risk assessment and high predictive performance. First, a multi-task sequence reconstruction strategy trains a shared denoising network on interpolation and extrapolation tasks to handle irregular follow-up sampling and learn robust latent trajectories. Second, an LLM-driven "linguistic compass" is introduced for clinical plausibility sampling: generated feature candidates are quantized, tokenized, and scored by a fine-tuned language model conditioned on expected structural biomarkers, guiding autoregressive generation toward realistic disease patterns. Experiments on ADNI and AIBL cohorts show that MCI-Diff outperforms state-of-the-art baselines, improving early conversion accuracy by 5-12%.

cs.LG↗

arXiv2025

Mobile-Agent-V: A Video-Guided Approach for Effortless and Efficient Operational Knowledge Injection in Mobile Automation

The exponential rise in mobile device usage necessitates streamlined automation for effective task management, yet many AI frameworks fall short due to inadequate operational expertise. While manually written knowledge can bridge this gap, it is often burdensome and inefficient. We introduce Mobile-Agent-V, an innovative framework that utilizes video as a guiding tool to effortlessly and efficiently inject operational knowledge into mobile automation processes. By deriving knowledge directly from video content, Mobile-Agent-V eliminates manual intervention, significantly reducing the effort and time required for knowledge acquisition. To rigorously evaluate this approach, we propose Mobile-Knowledge, a benchmark tailored to assess the impact of external knowledge on mobile agent performance. Our experimental findings demonstrate that Mobile-Agent-V enhances performance by 36% compared to existing methods, underscoring its effortless and efficient advantages in mobile automation.

cs.CL↗

arXiv2025

Mobile-Agent-V: A Video-Guided Approach for Effortless and Efficient Operational Knowledge Injection in Mobile Automation

cs.AI↗

arXiv2025

MirrorShield: Towards Universal Defense Against Jailbreaks via Entropy-Guided Mirror Crafting

Defending large language models (LLMs) against jailbreak attacks is crucial for ensuring their safe deployment. Existing defense strategies typically rely on predefined static criteria to differentiate between harmful and benign prompts. However, such rigid rules fail to accommodate the inherent complexity and dynamic nature of real-world jailbreak attacks. In this paper, we focus on the novel challenge of universal defense against diverse jailbreaks. We propose a new concept ``mirror'', which is a dynamically generated prompt that reflects the syntactic structure of the input while ensuring semantic safety. The discrepancies between input prompts and their corresponding mirrors serve as guiding principles for defense. A novel defense model, MirrorShield, is further proposed to detect and calibrate risky inputs based on the crafted mirrors. Evaluated on multiple benchmark datasets and compared against ten state-of-the-art attack methods, MirrorShield demonstrates superior defense performance and promising generalization capabilities.

cs.CR↗

arXiv2025

Unconventional domain tessellations in moiré-of-moiré lattices

Imposing incommensurable periodicity on the periodic atomic lattice can lead to complex structural phases consisting of locally periodic structure bounded by topological defects. Twisted trilayer graphene (TTG) is an ideal material platform to study the interplay between different atomic periodicities, which can be tuned by twist angles between the layers, leading to moiré-of-moiré lattices. Interlayer and intralayer interactions between two interfaces in TTG transform this moiré-of-moiré lattice into an intricate network of domain structures at small twist angles, which can harbor exotic electronic behaviors. Here we report a complete structural phase diagram of TTG with atomic scale lattice reconstruction. Using transmission electron microscopy combined with a new interatomic potential simulation, we show several large-scale moiré lattices, including triangular, kagome, and a corner-shared hexagram-shaped domain pattern. Each domain is bounded by a two-dimensional network of domain wall lattices. In the limit of small twist angles, two competing structural orders-rhombohedral and Bernal stackings-with a slight energy difference, cause unconventional lattice reconstruction with spontaneous symmetry breaking and nematic instability, highlighting the importance of long-range interlayer interactions across entire van der Waals layers. The diverse tessellation of distinct domains, whose topological network can be tuned by the adjustment of the twist angles, establishes TTG as a platform for exploring the interplay between emerging quantum properties and controllable nontrivial lattices.

cond-mat.mtrl-sci↗

arXiv2025

Hoist with His Own Petard: Inducing Guardrails to Facilitate Denial-of-Service Attacks on Retrieval-Augmented Generation of LLMs

Retrieval-Augmented Generation (RAG) integrates Large Language Models (LLMs) with external knowledge bases, improving output quality while introducing new security risks. Existing studies on RAG vulnerabilities typically focus on exploiting the retrieval mechanism to inject erroneous knowledge or malicious texts, inducing incorrect outputs. However, these approaches overlook critical weaknesses within LLMs, leaving important attack vectors unexplored and limiting the scope and efficiency of attacks. In this paper, we uncover a novel vulnerability: the safety guardrails of LLMs, while designed for protection, can also be exploited as an attack vector by adversaries. Building on this vulnerability, we propose MutedRAG, a novel denial-of-service attack that reversely leverages the guardrails of LLMs to undermine the availability of RAG systems. By injecting minimalistic jailbreak texts, such as "\textit{How to build a bomb}", into the knowledge base, MutedRAG intentionally triggers the LLM's safety guardrails, causing the system to reject legitimate queries. Besides, due to the high sensitivity of guardrails, a single jailbreak sample can affect multiple queries, effectively amplifying the efficiency of attacks while reducing their costs. Experimental results on three datasets demonstrate that MutedRAG achieves an attack success rate exceeding 60% in many scenarios, requiring only less than one malicious text to each target query on average. In addition, we evaluate potential defense strategies against MutedRAG, finding that some of current mechanisms are insufficient to mitigate this threat, underscoring the urgent need for more robust solutions.

cs.CR↗

arXiv2025

Analyzing the Role of the DSO in Electricity Trading of VPPs via a Stackelberg Game Model

The increasing penetration of distributed energy resources has sparked interests in participating in power markets. Here, we consider two settings where Virtual Power Plants (VPPs) with some flexible resources participate in the electricity trading, either directly in the wholesale electricity market, or interfaced by the Distribution System Operator (DSO) who is the transaction organizer. In order to study the role of DSO as a stakeholder, a Stackelberg game is represented via a bi-level model: the DSO maximizes profits at the upper level, while the VPPs minimize operating costs at the lower level. To solve this problem, the Karush-Kuhn-Tucker conditions of lower level is deduced to achieve a single-level problem. The results show that the role of the DSO as an intermediary agent leads to a decrease in operating costs of the VPPs by organizing lower-level trading, while making a profit for itself. However, this result comes with interests loss of the wholesale market, implying that stakeholders in the market need to abide by regulatory constraints.

eess.SY↗

arXiv2025

High-Dimensional Evolutionary Algorithm Based Design of Semi-Adder

Facing the physical limitations and energy consumption bottlenecks of traditional electronic devices, we propose an innovative design framework integrating evolutionary algorithms and metasurface technology, aiming to achieve intelligent inverse design of photonic devices. Based on a constructed high-dimensional evolutionary algorithm framework, a four-layer metasurface cascade regulation system was developed to realize the full optical physical expression of half-adder logic functions. This algorithm enables global optimization of 10000 unit parameters and can be extended to the design of more complex functional devices,thereby promoting goal-oriented and functional customization development

physics.optics↗

arXiv2025

Realization of a Dirac-vortex topological photonic crystal fiber

Photonic crystal fibers (PCFs) that trap and guide light using photonic bandgaps have revolutionized modern optics with enormous scientific innovations and technological applications spanning many disciplines. Recently, inspired by the discovery of topological phases of matter, Dirac-vortex topological PCFs have been theoretically proposed with intriguing topological properties and unprecedented opportunities in optical fiber communications. However, due to the substantial challenges of fabrication and characterization, experimental demonstration of Dirac-vortex topological PCFs has thus far remained elusive. Here, we report the experimental realization of a Dirac-vortex topological PCF using the standard stack-and-draw fabrication process with silica glass capillaries. Moreover, we experimentally observe that Dirac-vortex single-polarization single-mode bounds to and propagates along the fiber core in the full communication window (1260-1675nm). Our study pushes the research frontier of PCFs and provides a new avenue to enhance their performance and functionality further.

physics.optics↗

arXiv2025

Lattice dynamics and phonon dispersion of van der Waals layered ferromagnet Fe3GaTe2

Despite the tremendous progress in spintronic studies of van der Waals (vdW) room-temperature ferromagnet Fe3GaTe2, much less effort has been spent on its lattice dynamics and possible interaction with spintronic degrees of freedom. In this work, by combining Raman spectroscopy in a wide range of pressure (atmospheric pressure~19.5 GPa) and temperature (80~690 K) with first-principles calculation, we systematically studied the lattice dynamics and phonon dispersion of Fe3GaTe2. Our results show that the phonon energies of Fe3GaTe2 located at 126.0 cm-1 and 143.5 cm-1 originate from the anharmonic and harmonic vibration modes, respectively. Furthermore, the first room-temperature spin-phonon coupling in vdW ferromagnet is observed with strength of ~0.81 cm-1 at 300 K, by identifying Raman anomalies in both phonon energy and full width at half maximum (FWHM) of below Curie temperature of Fe3GaTe2. Our findings are valuable for fundamental and applied studies of vdW materials under variable conditions.

cond-mat.mtrl-sci↗

arXiv2025

Composite Nonlinear Trajectory Tracking Control of Co-Driving Vehicles Using Self-Triggered Adaptive Dynamic Programming

This article presents a composite nonlinear feedback (CNF) control method using self-triggered (ST) adaptive dynamic programming (ADP) algorithm in a human-machine shared steering framework. For the overall system dynamics, a two-degrees-of-freedom (2-DOF) vehicle model is established and a two-point preview driver model is adopted. A dynamic authority allocation strategy based on cooperation level is proposed to combine the steering input of the human driver and the automatic controller. To make further improvements in the controller design, three main contributions are put forward. Firstly, the CNF controller is designed for trajectory tracking control with refined transient performance. Besides, the self-triggered rule is applied such that the system will update in discrete times to save computing resources and increase efficiency. Moreover, by introducing the data-based ADP algorithm, the optimal control problem can be solved through iteration using system input and output information, reducing the need for accurate knowledge of system dynamics. The effectiveness of the proposed control method is validated through Carsim-Simulink co-simulations in diverse driving scenarios.

eess.SY↗

arXiv2025

Meta Flow Matching: Integrating Vector Fields on the Wasserstein Manifold

Numerous biological and physical processes can be modeled as systems of interacting entities evolving continuously over time, e.g. the dynamics of communicating cells or physical particles. Learning the dynamics of such systems is essential for predicting the temporal evolution of populations across novel samples and unseen environments. Flow-based models allow for learning these dynamics at the population level - they model the evolution of the entire distribution of samples. However, current flow-based models are limited to a single initial population and a set of predefined conditions which describe different dynamics. We argue that multiple processes in natural sciences have to be represented as vector fields on the Wasserstein manifold of probability densities. That is, the change of the population at any moment in time depends on the population itself due to the interactions between samples. In particular, this is crucial for personalized medicine where the development of diseases and their respective treatment response depend on the microenvironment of cells specific to each patient. We propose Meta Flow Matching (MFM), a practical approach to integrate along these vector fields on the Wasserstein manifold by amortizing the flow model over the initial populations. Namely, we embed the population of samples using a Graph Neural Network (GNN) and use these embeddings to train a Flow Matching model. This gives MFM the ability to generalize over the initial distributions, unlike previously proposed methods. We demonstrate the ability of MFM to improve the prediction of individual treatment responses on a large-scale multi-patient single-cell drug screen dataset.

cs.LG↗

arXiv2025

Internal Heat and Energy Imbalance of Uranus

With its extreme axial tilt, radiant energy budget and internal heat of Uranus remain among the most intriguing mysteries of our Solar System. Here, we present the global radiant energy budget spanning a complete orbital period, revealing significant seasonal variations driven primarily by the highly variable solar flux. Despite these fluctuations, emitted thermal power consistently exceeds absorbed solar power, indicating a net energy loss and ongoing global cooling. Based on the seasonal variations of radiant energy budget, we determine a statistically significant internal heat flux. This finding resolves a long-standing debate over whether Uranus possesses internal heat. We also examine the energy budget of the weather layer by combining the internal heat with the radiant energies, revealing significant energy imbalances at both global and hemispheric scales. These global and hemispheric imbalances should be considered in theoretical and numerical models. The Uranus flagship mission, as recommended by the recent survey, will provide crucial observations to address more unresolved questions and advance our understanding of this enigmatic ice giant.

astro-ph.EP↗

arXiv2025

JailBench: A Comprehensive Chinese Security Assessment Benchmark for Large Language Models

Large language models (LLMs) have demonstrated remarkable capabilities across various applications, highlighting the urgent need for comprehensive safety evaluations. In particular, the enhanced Chinese language proficiency of LLMs, combined with the unique characteristics and complexity of Chinese expressions, has driven the emergence of Chinese-specific benchmarks for safety assessment. However, these benchmarks generally fall short in effectively exposing LLM safety vulnerabilities. To address the gap, we introduce JailBench, the first comprehensive Chinese benchmark for evaluating deep-seated vulnerabilities in LLMs, featuring a refined hierarchical safety taxonomy tailored to the Chinese context. To improve generation efficiency, we employ a novel Automatic Jailbreak Prompt Engineer (AJPE) framework for JailBench construction, which incorporates jailbreak techniques to enhance assessing effectiveness and leverages LLMs to automatically scale up the dataset through context-learning. The proposed JailBench is extensively evaluated over 13 mainstream LLMs and achieves the highest attack success rate against ChatGPT compared to existing Chinese benchmarks, underscoring its efficacy in identifying latent vulnerabilities in LLMs, as well as illustrating the substantial room for improvement in the security and trustworthiness of LLMs within the Chinese context. Our benchmark is publicly available at https://github.com/STAIR-BUPT/JailBench.

cs.CL↗

arXiv2025

PC-Agent: A Hierarchical Multi-Agent Collaboration Framework for Complex Task Automation on PC

In the field of MLLM-based GUI agents, compared to smartphones, the PC scenario not only features a more complex interactive environment, but also involves more intricate intra- and inter-app workflows. To address these issues, we propose a hierarchical agent framework named PC-Agent. Specifically, from the perception perspective, we devise an Active Perception Module (APM) to overcome the inadequate abilities of current MLLMs in perceiving screenshot content. From the decision-making perspective, to handle complex user instructions and interdependent subtasks more effectively, we propose a hierarchical multi-agent collaboration architecture that decomposes decision-making processes into Instruction-Subtask-Action levels. Within this architecture, three agents (i.e., Manager, Progress and Decision) are set up for instruction decomposition, progress tracking and step-by-step decision-making respectively. Additionally, a Reflection agent is adopted to enable timely bottom-up error feedback and adjustment. We also introduce a new benchmark PC-Eval with 25 real-world complex instructions. Empirical results on PC-Eval show that our PC-Agent achieves a 32% absolute improvement of task success rate over previous state-of-the-art methods. The code is available at https://github.com/X-PLUG/MobileAgent/tree/main/PC-Agent.

cs.CV↗