arXiv subjects

Rui Zhang

Publications and source records attributed to Rui Zhang.

At least 361 records · Page 20Linked to original sources

arXiv2025

Sycophancy in Vision-Language Models: A Systematic Analysis and an Inference-Time Mitigation Framework

Large Vision-Language Models (LVLMs) have shown significant capability in vision-language understanding. However, one critical issue that persists in these models is sycophancy, where models are unduly influenced by leading or deceptive prompts, resulting in biased outputs and hallucinations. Despite the rapid development of LVLMs, evaluating and mitigating sycophancy remains largely under-explored. In this work, we fill this gap by systematically analyzing sycophancy across multiple vision-language benchmarks and propose an inference-time mitigation framework. We curate leading queries and quantify the susceptibility of state-of-the-art LVLMs to prompt-induced bias, revealing consistent performance degradation and instability across models and tasks. Our analysis further uncovers model-specific behavioral traits, such as sentiment sensitivity and prediction polarity shifts under sycophancy. To mitigate these issues, we propose a training-free, model-agnostic framework that operates entirely at inference time. Our approach first employs a query neutralizer, leveraging an language model to suppress implicit sycophantic bias in user queries. We then introduce a sycophancy-aware contrastive decoding mechanism that dynamically recalibrates token-level output distributions by contrasting responses to neutralized and leading queries. Finally, an adaptive logits refinement module further modifies the contrasted logits by integrating both a adaptive plausibility filter and query sentiment scaler, ensuring coherent and robust generation. Extensive experiments demonstrate that this framework effectively mitigates sycophancy across all evaluated models, while maintaining performance on neutral prompts. Our results suggest that sycophancy in LVLMs is a general and urgent challenge, and that inference-time strategies offer a promising path toward trustworthy multimodal reasoning.

cs.AI↗

arXiv2025

Large Language Models for Disease Diagnosis: A Scoping Review

Automatic disease diagnosis has become increasingly valuable in clinical practice. The advent of large language models (LLMs) has catalyzed a paradigm shift in artificial intelligence, with growing evidence supporting the efficacy of LLMs in diagnostic tasks. Despite the increasing attention in this field, a holistic view is still lacking. Many critical aspects remain unclear, such as the diseases and clinical data to which LLMs have been applied, the LLM techniques employed, and the evaluation methods used. In this article, we perform a comprehensive review of LLM-based methods for disease diagnosis. Our review examines the existing literature across various dimensions, including disease types and associated clinical specialties, clinical data, LLM techniques, and evaluation methods. Additionally, we offer recommendations for applying and evaluating LLMs for diagnostic tasks. Furthermore, we assess the limitations of current research and discuss future directions. To our knowledge, this is the first comprehensive review for LLM-based disease diagnosis.

cs.CL↗

arXiv2025

Omnidirectionally manipulated skyrmions in an orientationally chiral system

Skyrmions, originally from condensed matter physics, have been widely explored in various physical systems, including soft matter. A crucial challenge in manipulating topological solitary waves like skyrmions is controlling their flow on demand. Here, we control the arbitrary moving direction of skyrmions in a chiral liquid crystal system by adjusting the bias of the applied alternate current electric field. Specifically, the velocity, including both moving direction and speed can be continuously changed. The motion control of skyrmions originates from the symmetry breaking of the topological structure induced by flexoelectric-polarization effect. The omnidirectional control of topological solitons opens new avenues in light-steering and racetrack memories.

cond-mat.soft↗

arXiv2025

SpaceTrack-TimeSeries: Time Series Dataset towards Satellite Orbit Analysis

With the rapid advancement of aerospace technology and the large-scale deployment of low Earth orbit (LEO) satellite constellations, the challenges facing astronomical observations and deep space exploration have become increasingly pronounced. As a result, the demand for high-precision orbital data on space objects-along with comprehensive analyses of satellite positioning, constellation configurations, and deep space satellite dynamics-has grown more urgent. However, there remains a notable lack of publicly accessible, real-world datasets to support research in areas such as space object maneuver behavior prediction and collision risk assessment. This study seeks to address this gap by collecting and curating a representative dataset of maneuvering behavior from Starlink satellites. The dataset integrates Two-Line Element (TLE) catalog data with corresponding high-precision ephemeris data, thereby enabling a more realistic and multidimensional modeling of space object behavior. It provides valuable insights into practical deployment of maneuver detection methods and the evaluation of collision risks in increasingly congested orbital environments.

astro-ph.EP↗

arXiv2025

Locally Differentially Private Frequency Estimation via Joint Randomized Response

Local Differential Privacy (LDP) has been widely recognized as a powerful tool for providing a strong theoretical guarantee of data privacy to data contributors against an untrusted data collector. Under a typical LDP scheme, each data contributor independently randomly perturbs their data before submitting them to the data collector, which in turn infers valuable statistics about the original data from received perturbed data. Common to existing LDP mechanisms is an inherent trade-off between the level of privacy protection and data utility in the sense that strong data privacy often comes at the cost of reduced data utility. Frequency estimation based on Randomized Response (RR) is a fundamental building block of many LDP mechanisms. In this paper, we propose a novel Joint Randomized Response (JRR) mechanism based on correlated data perturbations to achieve locally differentially private frequency estimation. JRR divides data contributors into disjoint groups of two members and lets those in the same group jointly perturb their binary data to improve frequency-estimation accuracy and achieve the same level of data privacy by hiding the group membership information in contrast to the classical RR mechanism. Theoretical analysis and detailed simulation studies using both real and synthetic datasets show that JRR achieves the same level of data privacy as the classical RR mechanism while improving the frequency-estimation accuracy in the overwhelming majority of the cases by up to two orders of magnitude.

cs.CR↗

arXiv2025

Deep Sparse Latent Feature Models for Knowledge Graph Completion

Recent advances in knowledge graph completion (KGC) have emphasized text-based approaches to navigate the inherent complexities of large-scale knowledge graphs (KGs). While these methods have achieved notable progress, they frequently struggle to fully incorporate the global structural properties of the graph. Stochastic blockmodels (SBMs), especially the latent feature relational model (LFRM), offer robust probabilistic frameworks for identifying latent community structures and improving link prediction. This paper presents a novel probabilistic KGC framework utilizing sparse latent feature models, optimized via a deep variational autoencoder (VAE). Our proposed method dynamically integrates global clustering information with local textual features to effectively complete missing triples, while also providing enhanced interpretability of the underlying latent structures. Extensive experiments on four benchmark datasets with varying scales demonstrate the significant performance gains achieved by our method.

cs.CL↗

arXiv2025

Enhance Multimodal Consistency and Coherence for Text-Image Plan Generation

People get informed of a daily task plan through diverse media involving both texts and images. However, most prior research only focuses on LLM's capability of textual plan generation. The potential of large-scale models in providing text-image plans remains understudied. Generating high-quality text-image plans faces two main challenges: ensuring consistent alignment between two modalities and keeping coherence among visual steps. To address these challenges, we propose a novel framework that generates and refines text-image plans step-by-step. At each iteration, our framework (1) drafts the next textual step based on the prediction history; (2) edits the last visual step to obtain the next one; (3) extracts PDDL-like visual information; and (4) refines the draft with the extracted visual information. The textual and visual step produced in stage (4) and (2) will then serve as inputs for the next iteration. Our approach offers a plug-and-play improvement to various backbone models, such as Mistral-7B, Gemini-1.5, and GPT-4o. To evaluate the effectiveness of our approach, we collect a new benchmark consisting of 1,100 tasks and their text-image pair solutions covering 11 daily topics. We also design and validate a new set of metrics to evaluate the multimodal consistency and coherence in text-image plans. Extensive experiment results show the effectiveness of our approach on a range of backbone models against competitive baselines. Our code and data are available at https://github.com/psunlpgroup/MPlanner.

cs.CV↗

arXiv2025

Accelerating Causal Network Discovery of Alzheimer Disease Biomarkers via Scientific Literature-based Retrieval Augmented Generation

The causal relationships between biomarkers are essential for disease diagnosis and medical treatment planning. One notable application is Alzheimer's disease (AD) diagnosis, where certain biomarkers may influence the presence of others, enabling early detection, precise disease staging, targeted treatments, and improved monitoring of disease progression. However, understanding these causal relationships is complex and requires extensive research. Constructing a comprehensive causal network of biomarkers demands significant effort from human experts, who must analyze a vast number of research papers, and have bias in understanding diseases' biomarkers and their relation. This raises an important question: Can advanced large language models (LLMs), such as those utilizing retrieval-augmented generation (RAG), assist in building causal networks of biomarkers for further medical analysis? To explore this, we collected 200 AD-related research papers published over the past 25 years and then integrated scientific literature with RAG to extract AD biomarkers and generate causal relations among them. Given the high-risk nature of the medical diagnosis, we applied uncertainty estimation to assess the reliability of the generated causal edges and examined the faithfulness and scientificness of LLM reasoning using both automatic and human evaluation. We find that RAG enhances the ability of LLMs to generate more accurate causal networks from scientific papers. However, the overall performance of LLMs in identifying causal relations of AD biomarkers is still limited. We hope this study will inspire further foundational research on AI-driven analysis of AD biomarkers causal network discovery.

cs.IR↗

arXiv2025

Improving Fairness of Large Language Models in Multi-document Summarization

Fairness in multi-document summarization (MDS) is crucial for providing comprehensive views across documents with diverse social attribute values, which can significantly impact decision-making. For example, a summarization system that tends to overrepresent negative reviews of products can mislead customers into disregarding good products. Previous works measure fairness in MDS at two levels: summary-level and corpus-level. While summary-level fairness focuses on individual summaries, corpus-level fairness focuses on a corpus of summaries. Recent methods primarily focus on summary-level fairness. We propose FairPO, a preference tuning method that focuses on both summary-level and corpus-level fairness in MDS. To improve summary-level fairness, we propose to generate preference pairs by perturbing document sets. To improve corpus-level fairness, we propose fairness-aware preference tuning by dynamically adjusting the weights of preference pairs. Our experiments show that FairPO outperforms strong baselines while maintaining the critical qualities of summaries. The code is available at https://github.com/leehaoyuan/coverage_fairnes.

cs.CL↗

arXiv2025

Revisiting the electron affinity of selenium

The electron affinity (EA) of atomic selenium, previously established as 16,297.276(9) cm-1 based on the laser photodetachment microscopy (LPM) measurements in 2012, exhibited a significant deviation from other earlier experimental values, yet it remained the accepted reference standard for over a decade. In this letter, we re-examined the EA of Se using the slow-electron velocity-map imaging method and revealed a substantial deviation in the LPM result. Measurements for the different isotopes of Se and the energy-level splitting of the neutral Se atom's 3P2 - 3P1 further verified the accuracy and robustness of our SEVI method. Based on these experimental evidences, we recommended a revised EA(Se) value of 16,297.78(4) cm-1, which is in excellent agreement with the previous laser photodetachment threshold (LPT) experimental results.

physics.atom-ph↗

arXiv2025

The Freight Multimodal Transport Problem with Buses and Drones: An Integrated Approach for Last-Mile Delivery

This paper proposes a novel freight multimodal transport problem with buses and drones, where buses are responsible for transporting parcels to lockers at bus stops for storage, while drones are used to deliver each parcel from the locker to the corresponding customer. The integrated bus-drone system synergistically expands drone service coverage using the bus network to ensure efficient final delivery. Minimizing the total operational costs while satisfying customer demands necessitates the joint optimization of parcel assignments and drone flights. We model the problem into a compact mixed-integer linear programming formulation and propose an integer programming formulation with exponentially many variables. To address real-world scale instances, we propose a Branch-Price-and-Benders-Cut algorithm for this non-deterministic polynomial-time (NP)-hard problem. This algorithm, integrating column generation and Benders decomposition within a Branch-and-Bound framework, is developed to obtain optimal or near-optimal solutions. Additionally, we introduce algorithmic enhancements aimed at accelerating the convergence of the algorithm. Computational experiments on instances generated from real-world bus data demonstrate that the proposed algorithms outperform CPLEX regarding both efficiency and solution quality. Moreover, our approaches can lead to over 6% cost savings compared to situations where we determine parcel assignments and drone flights sequentially. We evaluate the environmental advantages of integrating buses and drones, study the impact of different cost parameters in the system, and investigate the impact of the parcel locker configuration on performance. These findings provide valuable managerial insights for urban logistics managers, highlighting the potential of the integrated bus-drone system to improve traditional last-mile delivery.

cs.DM↗

arXiv2025

TaxoAdapt: Aligning LLM-Based Multidimensional Taxonomy Construction to Evolving Research Corpora

The rapid evolution of scientific fields introduces challenges in organizing and retrieving scientific literature. While expert-curated taxonomies have traditionally addressed this need, the process is time-consuming and expensive. Furthermore, recent automatic taxonomy construction methods either (1) over-rely on a specific corpus, sacrificing generalizability, or (2) depend heavily on the general knowledge of large language models (LLMs) contained within their pre-training datasets, often overlooking the dynamic nature of evolving scientific domains. Additionally, these approaches fail to account for the multi-faceted nature of scientific literature, where a single research paper may contribute to multiple dimensions (e.g., methodology, new tasks, evaluation metrics, benchmarks). To address these gaps, we propose TaxoAdapt, a framework that dynamically adapts an LLM-generated taxonomy to a given corpus across multiple dimensions. TaxoAdapt performs iterative hierarchical classification, expanding both the taxonomy width and depth based on corpus' topical distribution. We demonstrate its state-of-the-art performance across a diverse set of computer science conferences over the years to showcase its ability to structure and capture the evolution of scientific fields. As a multidimensional method, TaxoAdapt generates taxonomies that are 26.51% more granularity-preserving and 50.41% more coherent than the most competitive baselines judged by LLMs.

cs.CL↗

arXiv2025

Unified model for non-Abelian braiding of Majorana and Dirac fermion zero modes

Majorana zero modes (MZMs) are the most intensively studied non-Abelian anyons. The Dirac fermion zero modes in topological insulators, which are symmetry-protected doubling of MZMs under fermion number conservation, offer an alternative approach to explore non-Abelian anyons. However, a unified model that elucidates the braiding statistics of these types of topological zero modes remains absent. We show that the minimal Kitaev chain model beyond fine-tuning regime provides a unified characterization of the non-Abelian statistics of both MZMs and Dirac fermion zero modes in different parameter regimes. In particular, we introduce a minimal tri-junction setting based on the minimal Kitaev chain model and show it facilitates the unified scheme of braiding Dirac fermion zero modes, as well as the MZMs in the assistance of a Dirac mode. This unified minimal model provides deeper insights into non-Abelian statistics, demonstrating that the non-Abelian braiding of MZMs can be continuously extended to encompass Dirac fermion zero modes. The minimal Kitaev chain has been realized in coupled quantum dots [Nature 614, 445 (2023)]. Our extension, which demonstrates novel nontrivial phases with non-Abelian MZM pairs and Dirac zero modes emerging in the broader parameter regimes without fine-tuning, expands the accessible experimental parameter space and enhances the feasibility of observing non-Abelian statistics in the minimal Kitaev chain model.

cond-mat.mes-hall↗

arXiv2025

Learning Robust Heterogeneous Graph Representations via Contrastive-Reconstruction under Sparse Semantics

In graph self-supervised learning, masked autoencoders (MAE) and contrastive learning (CL) are two prominent paradigms. MAE focuses on reconstructing masked elements, while CL maximizes similarity between augmented graph views. Recent studies highlight their complementarity: MAE excels at local feature capture, and CL at global information extraction. Hybrid frameworks for homogeneous graphs have been proposed, but face challenges in designing shared encoders to meet the semantic requirements of both tasks. In semantically sparse scenarios, CL struggles with view construction, and gradient imbalance between positive and negative samples persists. This paper introduces HetCRF, a novel dual-channel self-supervised learning framework for heterogeneous graphs. HetCRF uses a two-stage aggregation strategy to adapt embedding semantics, making it compatible with both MAE and CL. To address semantic sparsity, it enhances encoder output for view construction instead of relying on raw features, improving efficiency. Two positive sample augmentation strategies are also proposed to balance gradient contributions. Node classification experiments on four real-world heterogeneous graph datasets demonstrate that HetCRF outperforms state-of-the-art baselines. On datasets with missing node features, such as Aminer and Freebase, at a 40% label rate in node classification, HetCRF improves the Macro-F1 score by 2.75% and 2.2% respectively compared to the second-best baseline, validating its effectiveness and superiority.

cs.LG↗

arXiv2025

IMPA-HGAE:Intra-Meta-Path Augmented Heterogeneous Graph Autoencoder

Self-supervised learning (SSL) methods have been increasingly applied to diverse downstream tasks due to their superior generalization capabilities and low annotation costs. However, most existing heterogeneous graph SSL models convert heterogeneous graphs into homogeneous ones via meta-paths for training, which only leverage information from nodes at both ends of meta-paths while underutilizing the heterogeneous node information along the meta-paths. To address this limitation, this paper proposes a novel framework named IMPA-HGAE to enhance target node embeddings by fully exploiting internal node information along meta-paths. Experimental results validate that IMPA-HGAE achieves superior performance on heterogeneous datasets. Furthermore, this paper introduce innovative masking strategies to strengthen the representational capacity of generative SSL models on heterogeneous graph data. Additionally, this paper discuss the interpretability of the proposed method and potential future directions for generative self-supervised learning in heterogeneous graphs. This work provides insights into leveraging meta-path-guided structural semantics for robust representation learning in complex graph scenarios.

cs.LG↗

arXiv2025

Trajectory Entropy: Modeling Game State Stability from Multimodality Trajectory Prediction

Complex interactions among agents present a significant challenge for autonomous driving in real-world scenarios. Recently, a promising approach has emerged, which formulates the interactions of agents as a level-k game framework. It effectively decouples agent policies by hierarchical game levels. However, this framework ignores both the varying driving complexities among agents and the dynamic changes in agent states across game levels, instead treating them uniformly. Consequently, redundant and error-prone computations are introduced into this framework. To tackle the issue, this paper proposes a metric, termed as Trajectory Entropy, to reveal the game status of agents within the level-k game framework. The key insight stems from recognizing the inherit relationship between agent policy uncertainty and the associated driving complexity. Specifically, Trajectory Entropy extracts statistical signals representing uncertainty from the multimodality trajectory prediction results of agents in the game. Then, the signal-to-noise ratio of this signal is utilized to quantify the game status of agents. Based on the proposed Trajectory Entropy, we refine the current level-k game framework through a simple gating mechanism, significantly improving overall accuracy while reducing computational costs. Our method is evaluated on the Waymo and nuPlan datasets, in terms of trajectory prediction, open-loop and closed-loop planning tasks. The results demonstrate the state-of-the-art performance of our method, with precision improved by up to 19.89% for prediction and up to 16.48% for planning.

cs.AI↗

arXiv2025

QiMeng: Fully Automated Hardware and Software Design for Processor Chip

Processor chip design technology serves as a key frontier driving breakthroughs in computer science and related fields. With the rapid advancement of information technology, conventional design paradigms face three major challenges: the physical constraints of fabrication technologies, the escalating demands for design resources, and the increasing diversity of ecosystems. Automated processor chip design has emerged as a transformative solution to address these challenges. While recent breakthroughs in Artificial Intelligence (AI), particularly Large Language Models (LLMs) techniques, have opened new possibilities for fully automated processor chip design, substantial challenges remain in establishing domain-specific LLMs for processor chip design. In this paper, we propose QiMeng, a novel system for fully automated hardware and software design of processor chips. QiMeng comprises three hierarchical layers. In the bottom-layer, we construct a domain-specific Large Processor Chip Model (LPCM) that introduces novel designs in architecture, training, and inference, to address key challenges such as knowledge representation gap, data scarcity, correctness assurance, and enormous solution space. In the middle-layer, leveraging the LPCM's knowledge representation and inference capabilities, we develop the Hardware Design Agent and the Software Design Agent to automate the design of hardware and software for processor chips. Currently, several components of QiMeng have been completed and successfully applied in various top-layer applications, demonstrating significant advantages and providing a feasible solution for efficient, fully automated hardware/software design of processor chips. Future research will focus on integrating all components and performing iterative top-down and bottom-up design processes to establish a comprehensive QiMeng system.

cs.AR↗

arXiv2025

Statistical Channel Based Low-Complexity Rotation and Position Optimization for 6D Movable Antennas Enabled Wireless Communication

Six-dimensional movable antenna (6DMA) is a promising technology to fully exploit spatial variation in wireless channels by allowing flexible adjustment of three-dimensional (3D) positions and rotations of antennas at the transceiver. In this paper, we investigate the practical low-complexity design of 6DMA-enabled communication systems, including transmission protocol, statistical channel information (SCI) acquisition, and joint position and rotation optimization of 6DMA surfaces based on the SCI of users. Specifically, an orthogonal matching pursuit (OMP)-based algorithm is proposed for the estimation of SCI of users at all possible position-rotation pairs of 6DMA surfaces based on the channel measurements at a small subset of position-rotation pairs. Then, the average sum logarithmic rate of all users is maximized by jointly designing the positions and rotations of 6DMA surfaces based on their SCI acquired. Different from prior works on 6DMA which adopt alternating optimization to design 6DMA positions/rotations with iterations, we propose a new sequential optimization approach that first determines 6DMA rotations and then finds their feasible positions to realize the optimized rotations subject to practical antenna placement constraints. Simulation results show that the proposed sequential optimization significantly reduces the computational complexity of conventional alternating optimization, while achieving comparable communication performance. It is also shown that the proposed SCI-based 6DMA design can effectively enhance the communication throughput of wireless networks over existing fixed (position and rotation) antenna arrays, yet with a practically appealing low-complexity implementation.

cs.IT↗