arXiv subjects

Zheng Wang

Publications and source records attributed to Zheng Wang.

At least 163 records · Page 9Linked to original sources

arXiv2025

Detecting Corpus-Level Knowledge Inconsistencies in Wikipedia with Large Language Models

Wikipedia is the largest open knowledge corpus, widely used worldwide and serving as a key resource for training large language models (LLMs) and retrieval-augmented generation (RAG) systems. Ensuring its accuracy is therefore critical. But how accurate is Wikipedia, and how can we improve it? We focus on inconsistencies, a specific type of factual inaccuracy, and introduce the task of corpus-level inconsistency detection. We present CLAIRE, an agentic system that combines LLM reasoning with retrieval to surface potentially inconsistent claims along with contextual evidence for human review. In a user study with experienced Wikipedia editors, 87.5% reported higher confidence when using CLAIRE, and participants identified 64.7% more inconsistencies in the same amount of time. Combining CLAIRE with human annotation, we contribute WIKICOLLIDE, the first benchmark of real Wikipedia inconsistencies. Using random sampling with CLAIRE-assisted analysis, we find that at least 3.3% of English Wikipedia facts contradict another fact, with inconsistencies propagating into 7.3% of FEVEROUS and 4.0% of AmbigQA examples. Benchmarking strong baselines on this dataset reveals substantial headroom: the best fully automated system achieves an AUROC of only 75.1%. Our results show that contradictions are a measurable component of Wikipedia and that LLM-based systems like CLAIRE can provide a practical tool to help editors improve knowledge consistency at scale.

cs.CL↗

arXiv2025

Subjective Camera 1.0: Bridging Human Cognition and Visual Reconstruction through Sequence-Aware Sketch-Guided Diffusion

We introduce the concept of a subjective camera to reconstruct meaningful moments that physical cameras fail to capture. We propose Subjective Camera 1.0, a framework for reconstructing real-world scenes from readily accessible subjective readouts, i.e., textual descriptions and progressively drawn rough sketches. Built on optimization-based alignment of diffusion models, our approach avoids large-scale paired training data and mitigates generalization issues. To address the challenge of integrating multiple abstract concepts in real-world scenarios, we design a Sequence-Aware Sketch-Guided Diffusion framework with three loss terms for concept-wise sequential optimization, following the natural order of subjective readouts. Experiments on two datasets demonstrate that our method achieves state-of-the-art performance in image quality as well as spatial and semantic alignment with target scenes. User studies with 40 participants further confirm that our approach is consistently preferred. Our project page is at: subjective-camera.github.io

cs.CV↗

arXiv2025

Enhancing WSI-Based Survival Analysis with Report-Auxiliary Self-Distillation

Survival analysis based on Whole Slide Images (WSIs) is crucial for evaluating cancer prognosis, as they offer detailed microscopic information essential for predicting patient outcomes. However, traditional WSI-based survival analysis usually faces noisy features and limited data accessibility, hindering their ability to capture critical prognostic features effectively. Although pathology reports provide rich patient-specific information that could assist analysis, their potential to enhance WSI-based survival analysis remains largely unexplored. To this end, this paper proposes a novel Report-auxiliary self-distillation (Rasa) framework for WSI-based survival analysis. First, advanced large language models (LLMs) are utilized to extract fine-grained, WSI-relevant textual descriptions from original noisy pathology reports via a carefully designed task prompt. Next, a self-distillation-based pipeline is designed to filter out irrelevant or redundant WSI features for the student model under the guidance of the teacher model's textual knowledge. Finally, a risk-aware mix-up strategy is incorporated during the training of the student model to enhance both the quantity and diversity of the training data. Extensive experiments carried out on our collected data (CRC) and public data (TCGA-BRCA) demonstrate the superior effectiveness of Rasa against state-of-the-art methods. Our code is available at https://github.com/zhengwang9/Rasa.

cs.CV↗

arXiv2025

Network stochastic resonance under higher-order interactions

Although stochastic resonance phenomena are ubiquitous across various complex systems, the influence mechanisms of higher-order interactions remain elusive. Here, we address this gap by investigating stochastic resonance in coupled phase oscillators with triadic interactions on ring networks that feature periodic modulation of the triadic coupling strength and additive noise. Our analysis reveals that higher-order interactions create fundamentally different resonance landscapes through time-varying potential wells generated by periodic modulation of triadic coupling, enabling novel noise-enhanced processing mechanisms. Additionally, weaker pairwise coupling amplifies resonance effects, while moderate network connectivity appears optimal compared to extensive connections. Our findings establish fundamental principles for network stochastic resonance and provide insights for enhanced signal processing in complex networks.

nlin.AO↗

arXiv2025

AI Agentic Programming: A Survey of Techniques, Challenges, and Opportunities

AI agentic programming is an emerging paradigm where large language model (LLM)-based coding agents autonomously plan, execute, and interact with tools such as compilers, debuggers, and version control systems. Unlike conventional code generation, these agents decompose goals, coordinate multi-step processes, and adapt based on feedback, reshaping software development practices. This survey provides a timely review of the field, introducing a taxonomy of agent behaviors and system architectures and examining relevant techniques for planning, context management, tool integration, execution monitoring, and benchmarking datasets. We highlight challenges of this fast-moving field and discuss opportunities for building reliable, transparent, and collaborative coding agents.

cs.SE↗

arXiv2025

A New Dataset and Benchmark for Grounding Multimodal Misinformation

The proliferation of online misinformation videos poses serious societal risks. Current datasets and detection methods primarily target binary classification or single-modality localization based on post-processed data, lacking the interpretability needed to counter persuasive misinformation. In this paper, we introduce the task of Grounding Multimodal Misinformation (GroundMM), which verifies multimodal content and localizes misleading segments across modalities. We present the first real-world dataset for this task, GroundLie360, featuring a taxonomy of misinformation types, fine-grained annotations across text, speech, and visuals, and validation with Snopes evidence and annotator reasoning. We also propose a VLM-based, QA-driven baseline, FakeMark, using single- and cross-modal cues for effective detection and grounding. Our experiments highlight the challenges of this task and lay a foundation for explainable multimodal misinformation detection.

cs.SI↗

arXiv2025

MiniCPM4: Ultra-Efficient LLMs on End Devices

This paper introduces MiniCPM4, a highly efficient large language model (LLM) designed explicitly for end-side devices. We achieve this efficiency through systematic innovation in four key dimensions: model architecture, training data, training algorithms, and inference systems. Specifically, in terms of model architecture, we propose InfLLM v2, a trainable sparse attention mechanism that accelerates both prefilling and decoding phases for long-context processing. Regarding training data, we propose UltraClean, an efficient and accurate pre-training data filtering and generation strategy, and UltraChat v2, a comprehensive supervised fine-tuning dataset. These datasets enable satisfactory model performance to be achieved using just 8 trillion training tokens. Regarding training algorithms, we propose ModelTunnel v2 for efficient pre-training strategy search, and improve existing post-training methods by introducing chunk-wise rollout for load-balanced reinforcement learning and data-efficient tenary LLM, BitCPM. Regarding inference systems, we propose CPM.cu that integrates sparse attention, model quantization, and speculative sampling to achieve efficient prefilling and decoding. To meet diverse on-device requirements, MiniCPM4 is available in two versions, with 0.5B and 8B parameters, respectively. Furthermore, we construct a hybrid reasoning model, MiniCPM4.1, which can be used in both deep reasoning mode and non-reasoning mode. Evaluation results demonstrate that MiniCPM4 and MiniCPM4.1 outperform similar-sized open-source models across benchmarks, with the 8B variants showing significant speed improvements on long sequence understanding and generation.

cs.CL↗

arXiv2025

Multi-Type Context-Aware Conversational Recommender Systems via Mixture-of-Experts

Conversational recommender systems enable natural language conversations and thus lead to a more engaging and effective recommendation scenario. As the conversations for recommender systems usually contain limited contextual information, many existing conversational recommender systems incorporate external sources to enrich the contextual information. However, how to combine different types of contextual information is still a challenge. In this paper, we propose a multi-type context-aware conversational recommender system, called MCCRS, effectively fusing multi-type contextual information via mixture-of-experts to improve conversational recommender systems. MCCRS incorporates both structured information and unstructured information, including the structured knowledge graph, unstructured conversation history, and unstructured item reviews. It consists of several experts, with each expert specialized in a particular domain (i.e., one specific contextual information). Multiple experts are then coordinated by a ChairBot to generate the final results. Our proposed MCCRS model takes advantage of different contextual information and the specialization of different experts followed by a ChairBot breaks the model bottleneck on a single contextual information. Experimental results demonstrate that our proposed MCCRS method achieves significantly higher performance compared to existing baselines.

cs.CL↗

arXiv2025

TrajSV: A Trajectory-based Model for Sports Video Representations and Applications

Sports analytics has received significant attention from both academia and industry in recent years. Despite the growing interest and efforts in this field, several issues remain unresolved, including (1) data unavailability, (2) lack of an effective trajectory-based framework, and (3) requirement for sufficient supervision labels. In this paper, we present TrajSV, a trajectory-based framework that addresses various issues in existing studies. TrajSV comprises three components: data preprocessing, Clip Representation Network (CRNet), and Video Representation Network (VRNet). The data preprocessing module extracts player and ball trajectories from sports broadcast videos. CRNet utilizes a trajectory-enhanced Transformer module to learn clip representations based on these trajectories. Additionally, VRNet learns video representations by aggregating clip representations and visual features with an encoder-decoder architecture. Finally, a triple contrastive loss is introduced to optimize both video and clip representations in an unsupervised manner. The experiments are conducted on three broadcast video datasets to verify the effectiveness of TrajSV for three types of sports (i.e., soccer, basketball, and volleyball) with three downstream applications (i.e., sports video retrieval, action spotting, and video captioning). The results demonstrate that TrajSV achieves state-of-the-art performance in sports video retrieval, showcasing a nearly 70% improvement. It outperforms baselines in action spotting, achieving state-of-the-art results in 9 out of 17 action categories, and demonstrates a nearly 20% improvement in video captioning. Additionally, we introduce a deployed system along with the three applications based on TrajSV.

cs.CV↗

arXiv2025

MetaGuardian: Enhancing Voice Assistant Security through Advanced Acoustic Metamaterials

We present MetaGuardian, a voice assistant (VA) protection system based on acoustic metamaterials. MetaGuardian can be directly integrated into the enclosures of various smart devices, effectively defending against inaudible, adversarial and laser attacks without relying on additional software support or altering the underlying hardware, ensuring usability. To achieve this, MetaGuardian leverages the mutual impedance effects between metamaterial units to extend the signal filtering range to 16-40 kHz to effectively block wide-band inaudible attacks. Additionally, it adopts a carefully designed coiled space structure to precisely interfere with adversarial attacks while ensuring the normal functioning of VAs. Furthermore, MetaGuardian offers a universal structural design, allowing itself to be flexibly adapted to various smart devices, striking a balance between portability and protection effectiveness. In controled evaluation environments, MetaGuardian achieves a high defense success rate against various attack types, including adversarial, inaudible and laser attacks.

cs.SD↗

arXiv2025

New Constructions of Locally Perfect Nonlinear Functions and Their Application to Sequence Sets With Low Ambiguity Zone

Low Ambiguity Zone (LAZ) sequences play a pivotal role in modern integrated sensing and communication (ISAC) systems. Recently, Wang \textit{et al.} [arXiv:2501.11313] proposed a definition of locally perfect nonlinear functions (LPNFs) and constructed three classes of both periodic and aperiodic LAZ sequence sets with flexible parameters by applying such functions and interleaving techniques. Some of these LAZ sequence sets are asymptotically optimal with respect to the Ye-Zhou-Fan-Liu-Lei-Tang bounds under certain conditions. In this paper, we present constructions of three new classes of LPNFs with new parameters. Based on these LPNFs, we further propose a series of LAZ sequence sets that offer more flexible parameters. Furthermore, our results show that some of these classes are asymptotically optimal in both the periodic and aperiodic cases, respectively.

cs.IT↗

arXiv2025

Universally Unfiltered and Unseen:Input-Agnostic Multimodal Jailbreaks against Text-to-Image Model Safeguards

Various (text) prompt filters and (image) safety checkers have been implemented to mitigate the misuse of Text-to-Image (T2I) models in creating Not-Safe-For-Work (NSFW) content. In order to expose potential security vulnerabilities of such safeguards, multimodal jailbreaks have been studied. However, existing jailbreaks are limited to prompt-specific and image-specific perturbations, which suffer from poor scalability and time-consuming optimization. To address these limitations, we propose Universally Unfiltered and Unseen (U3)-Attack, a multimodal jailbreak attack method against T2I safeguards. Specifically, U3-Attack optimizes an adversarial patch on the image background to universally bypass safety checkers and optimizes a safe paraphrase set from a sensitive word to universally bypass prompt filters while eliminating redundant computations. Extensive experimental results demonstrate the superiority of our U3-Attack on both open-source and commercial T2I models. For example, on the commercial Runway-inpainting model with both prompt filter and safety checker, our U3-Attack achieves $~4\times$ higher success rates than the state-of-the-art multimodal jailbreak attack, MMA-Diffusion.

cs.CR↗

arXiv2025

Spectrum Sharing in Satellite-Terrestrial Integrated Networks: Frameworks, Approaches, and Opportunities

With the construction of low-earth orbit (LEO) satellite constellations, ubiquitous connectivity has been achieved. Terrestrial networks (TNs), such as cellular networks, are mainly deployed in specific urban areas and use licensed spectrum. However, in remote areas where terrestrial infrastructure is sparse, licensed spectrum bands are often underutilized. To accommodate the increasing communication needs, non-terrestrial networks (NTNs) can opportunistically access this idle spectrum to improve spectrum efficiency via spectrum sharing (SS). Therefore, bringing NTNs to a shared spectrum with TNs can improve network capacity under reasonable interference management. In satellite-terrestrial integrated networks (STINs), the comprehensive coverage of a satellite and the unbalanced communication resources of STINs make it challenging to manage mutual interference between NTN and TN effectively. This article presents the fundamentals and prospects of SS in STINs by introducing four SS frameworks, their potential application scenarios, and technical challenges. Furthermore, advanced SS approaches related to interference management in STINs and performance metrics of SS in STINs are introduced. Moreover, a preliminary performance evaluation showcases the potential for sharing the spectrum between NTN and TN. Finally, future research opportunities for SS in STINs are discussed.

eess.SP↗

arXiv2025

From Cluster Assumption to Graph Convolution: Graph-based Semi-Supervised Learning Revisited

Graph-based semi-supervised learning (GSSL) has long been a hot research topic. Traditional methods are generally shallow learners, based on the cluster assumption. Recently, graph convolutional networks (GCNs) have become the predominant techniques for their promising performance. In this paper, we theoretically discuss the relationship between these two types of methods in a unified optimization framework. One of the most intriguing findings is that, unlike traditional ones, typical GCNs may not jointly consider the graph structure and label information at each layer. Motivated by this, we further propose three simple but powerful graph convolution methods. The first is a supervised method OGC which guides the graph convolution process with labels. The others are two unsupervised methods: GGC and its multi-scale version GGCM, both aiming to preserve the graph structure information during the convolution process. Finally, we conduct extensive experiments to show the effectiveness of our methods. Code is available at https://github.com/zhengwang100/ogc_ggcm.

cs.LG↗

arXiv2025

Plant-Centric Metaverse: A Biocentric-Creation Framework for Ecological Art and Digital Symbiosis

Digital ecological art represents an emergent frontier where biological media converge with virtual environments. This study examines the paradigm shift from anthropocentric to plant-centered artistic narratives within the metaverse, contextualizing how digital platforms transform ecological expression. However, current frameworks fail to systematically guide artists in leveraging plant agency for digital symbiosis that transcends human-centered creation. We propose the Biocentric-Creation Transformation Ideology (BCTI) framework and validate it through multimodal case studies spanning bio-art, NFTs, and VR ecosystems (2013-2023). Our analysis reveals: (1) Metaverse ecosystems enable unprecedented plant-algorithm co-creation, with biological artworks increasing by 133% in premier archives (2020 vs 2013); (2) Digital symbiosis manifests through blockchain DAOs where plants govern human-plant collaborations; (3) Algorithmic photosynthesis in VR environments reshapes ecological aesthetics through real-time biodata translation. The BCTI framework advances ecological art theory by systematizing the transition from representation to plant-centered agency, offering artists a blueprint for post-anthropocene creation. This redefines environmental consciousness in virtual realms while establishing new protocols for cross-species digital collaboration.

cs.HC↗

arXiv2025

Network Embedding with Completely-imbalanced Labels

Network embedding, aiming to project a network into a low-dimensional space, is increasingly becoming a focus of network research. Semi-supervised network embedding takes advantage of labeled data, and has shown promising performance. However, existing semi-supervised methods would get unappealing results in the completely-imbalanced label setting where some classes have no labeled nodes at all. To alleviate this, we propose two novel semi-supervised network embedding methods. The first one is a shallow method named RSDNE. Specifically, to benefit from the completely-imbalanced labels, RSDNE guarantees both intra-class similarity and inter-class dissimilarity in an approximate way. The other method is RECT which is a new class of graph neural networks. Different from RSDNE, to benefit from the completely-imbalanced labels, RECT explores the class-semantic knowledge. This enables RECT to handle networks with node features and multi-label setting. Experimental results on several real-world datasets demonstrate the superiority of the proposed methods. Code is available at https://github.com/zhengwang100/RECT.

cs.LG↗

arXiv2025

Fine-grained Alignment of Large Language Models for General Medication Recommendation without Overprescription

Large language models (LLMs) holds significant promise in achieving general medication recommendation systems owing to their comprehensive interpretation of clinical notes and flexibility to medication encoding. We evaluated both general-purpose and medical-specific LLMs for medication recommendations, showing their unsatisfactory precision and severe overprescription. To address this, we introduce Language-Assisted Medication Recommendation, which tailors LLMs for medication recommendation in a medication-aware manner, improving the usage of clinical notes. Fine-tuning LLMs with this framework can outperform existing methods by more than 10% in internal validation and generalize across temporal and external validations. Furthermore, the model maintains high accuracy when encountering out-of-distribution medication.

cs.IR↗

arXiv2025

An Investigation into Value Misalignment in LLM-Generated Texts for Cultural Heritage

As Large Language Models (LLMs) become increasingly prevalent in tasks related to cultural heritage, such as generating descriptions of historical monuments, translating ancient texts, preserving oral traditions, and creating educational content, their ability to produce accurate and culturally aligned texts is being increasingly relied upon by users and researchers. However, cultural value misalignments may exist in generated texts, such as the misrepresentation of historical facts, the erosion of cultural identity, and the oversimplification of complex cultural narratives, which may lead to severe consequences. Therefore, investigating value misalignment in the context of LLM for cultural heritage is crucial for mitigating these risks, yet there has been a significant lack of systematic and comprehensive study and investigation in this area. To fill this gap, we systematically assess the reliability of LLMs in generating culturally aligned texts for cultural heritage-related tasks. We conduct a comprehensive evaluation by compiling an extensive set of 1066 query tasks covering 5 widely recognized categories with 17 aspects within the knowledge framework of cultural heritage across 5 open-source LLMs, and examine both the type and rate of cultural value misalignments in the generated texts. Using both automated and manual approaches, we effectively detect and analyze the cultural value misalignments in LLM-generated texts. Our findings are concerning: over 65% of the generated texts exhibit notable cultural misalignments, with certain tasks demonstrating almost complete misalignment with key cultural values. Beyond these findings, this paper introduces a benchmark dataset and a comprehensive evaluation workflow that can serve as a valuable resource for future research aimed at enhancing the cultural sensitivity and reliability of LLMs.

cs.CL↗