论文检索

Dahai Yu, Rongchao Xu, Dingyi Zhuang, Yuheng Bu, Shenhao Wang, Guang Wang

Energy usage prediction is important for various real-world applications, including grid management, infrastructure planning, and disaster response. Although a plethora of deep learning approaches have been proposed to perform this task, most of them either overlook the essential spatial correlations across households or fail to scale to individualized prediction, making them less effective for accurate fine-grained user-level prediction. In addition, due to the dynamic and uncertain nature of energy usage caused by various factors such as extreme weather events, quantifying uncertainty for reliable prediction is also significant, but it has not been fully explored in existing work. In this paper, we propose a unified framework called TrustEnergy for accurate and reliable user-level energy usage prediction. There are two key technical components in TrustEnergy, (i) a Hierarchical Spatiotemporal Representation module to efficiently capture both macro and micro energy usage patterns with a novel memory-augmented spatiotemporal graph neural network, and (ii) an innovative Sequential Conformalized Quantile Regression module to dynamically adjust uncertainty bounds to ensure valid prediction intervals over time, without making strong assumptions about the underlying data distribution. We implement and evaluate our TrustEnergy framework by working with an electricity provider in Florida, and the results show our TrustEnergy can achieve a 5.4% increase in prediction accuracy and 5.7% improvement in uncertainty quantification compared to state-of-the-art baselines.

Muhammad Hamza Yousuf, Jason Li, Sahar Vahdati, Raphael Theilen, Jakob Wittenstein, Jens Lehmann

Invasive mechanical ventilation (MV) is a life-sustaining therapy commonly used in the intensive care unit (ICU) for patients with severe and acute conditions. These patients frequently rely on MV for breathing. Given the high risk of death in such cases, optimal MV settings can reduce mortality, minimize ventilator-induced lung injury, shorten ICU stays, and ease the strain on healthcare resources. However, optimizing MV settings remains a complex and error-prone process due to patient-specific variability. While Offline Reinforcement Learning (RL) shows promise for optimizing MV settings, current methods struggle with the hybrid (continuous and discrete) nature of MV settings. Discretizing continuous settings leads to exponential growth in the action space, which limits the number of optimizable settings. Converting the predictions back to continuous can cause a distribution shift, compromising safety and performance. To address this challenge, in the IntelliLung project, we are developing an AI-based approach where we constrain the action space and employ factored action critics. This approach allows us to scale to six optimizable settings compared to 2-3 in previous studies. We adapt SOTA offline RL algorithms to operate directly on hybrid action spaces, avoiding the pitfalls of discretization. We also introduce a clinically grounded reward function based on ventilator-free days and physiological targets. Using multi-objective optimization for reward selection, we show that this leads to a more equitable consideration of all clinically relevant objectives. Notably, we develop a system in close collaboration with healthcare professionals that is aligned with real-world clinical objectives and designed with future deployment in mind.

Zonghai Yao, Talha Chafekar, Junda Wang, Shuo Han, Feiyun Ouyang, Junhui Qian, Lingxi Li, Hong Yu

Real-world adoption of closed-loop insulin delivery systems (CLIDS) in type 1 diabetes remains low, driven not by technical failure, but by diverse behavioral, psychosocial, and social barriers. We introduce ChatCLIDS, the first benchmark to rigorously evaluate LLM–driven persuasive dialogue for health behavior change. Our framework features a library of expert-validated virtual patients, each with clinically grounded, heterogeneous profiles and realistic adoption barriers, and simulates multi-turn interactions with nurse agents equipped with a diverse set of evidence-based persuasive strategies. ChatCLIDS uniquely supports longitudinal counseling and adversarial social influence scenarios, enabling robust, multi-dimensional evaluation. Our findings reveal that while larger and more reflective LLMs adapt strategies over time, all models struggle to overcome resistance, especially under realistic social pressure. These results highlight critical limitations of current LLMs for behavior change, and offer a high-fidelity, scalable testbed for advancing trustworthy persuasive AI in healthcare and beyond.

Yajie Yang, Yuqing Zhao, Xiaochao Xi, Yinan Zhu

Artificial Intelligence Generated Content (AIGC) assisting image production triggers controversy in journalism while attracting attention from media agencies. Key issues involve misinformation, authenticity, semantic fidelity, and interpretability. Most AIGC tools are opaque “black boxes,” hindering the dual demands of content accuracy and semantic alignment and creating ethical, sociotechnical, and trust dilemmas. This paper explores pathways for controllable image production in journalism’s special coverage and conducts two experiments with projects from China’s media agency: (1) Experiment 1 tests cross-platform adaptability via standardized prompts across three scenes, revealing disparities in semantic alignment, cultural specificity, and visual realism driven by training-corpus bias and platform-level filtering. (2) Experiment 2 builds a human-in-the-loop modular pipeline combining high-precision segmentation (SAM, GroundingDINO), semantic alignment (BrushNet), and style regulating (Style-LoRA, Prompt-to-Prompt), ensuring editorial fidelity through CLIP-based semantic scoring, NSFW/OCR/YOLO filtering, and verifiable content credentials. Traceable deployment preserves semantic representation. Consequently, we propose a human-AI collaboration mechanism for AIGC assisted image production in special coverage and recommend evaluating Character Identity Stability (CIS), Cultural Expression Accuracy (CEA), and User-Public Appropriateness (U-PA).

Qihao Yang, Xuelin Wang, Jiale Chen, Xuelian Dong, Yuxin Hao, Tianyong Hao

Language acquisition is vital to revealing the nature of human language intelligence and has recently emerged as a promising perspective for improving the interpretability of large language models (LLMs). However, it is ethically and practically infeasible to conduct experiments that require controlling human learners' language inputs. This poses challenges for the verifiability and scalability of language acquisition modeling, particularly in Chinese second language acquisition (SLA). While LLMs provide a controllable and reproducible alternative, a systematic benchmark to support phase-wise modeling and assessment is still lacking. To address these issues, we propose HSKBenchmark, the first benchmark for staged modeling and writing assessment of LLMs in Chinese SLA. The benchmark covers HSK levels 3 to 6, comprising authentic textbooks with 6.76M tokens, 16K synthetic instruction data, 30 test topics and a linguistically-grounded evaluation system. To simulate human acquisition trajectories, a curriculum-tuning framework is introduced, which trains LLMs in a progression from beginner to advanced proficiency levels. Since language production in writing is a key perspective for observing SLA development, an evaluation system is established to probe LLMs in writing, including the coverage of level-based grammar items, writing errors, lexical complexity, syntactic complexity, and holistic scoring. We also develop an HSKAgent fine-tuned on 10K compositions from Chinese second language learners to automate this evaluation system. Extensive experimental results demonstrate that HSKBenchmark not only models Chinese SLA effectively, but also serves as a reliable benchmark for dynamic writing assessment in LLMs. Our fine-tuned LLMs have writing performance on par with advanced human learners and exhibit human-like acquisition characteristics. The HSKBenchmark, HSKAgent, and checkpoints serve as foundational tools and resources, with the potential to pave the way for future research on language acquisition modeling and LLMs interpretability.

Luming Yang, Haoxian Liu, Siqing Li, Alper Yilmaz

Fine-grained action evaluation in medical vision faces unique challenges due to the unavailability of comprehensive datasets, stringent precision requirements, and insufficient spatiotemporal dynamic modeling of very rapid actions. To support development and evaluation, we introduce CPREval-6k, a multi-view, multi-label medical action benchmark containing 6,372 expert-annotated videos with 22 clinical labels. Using this dataset, we present GaussMedAct, a multivariate Gaussian encoding framework, to advance medical motion analysis through adaptive spatiotemporal representation learning. Multivariate Gaussian Representation projects the joint motions to a temporally scaled multi-dimensional space, and decomposes actions into adaptive 3D Gaussians that serve as tokens. These tokens preserve motion semantics through anisotropic covariance modeling while maintaining robustness to spatiotemporal noise. Hybrid Spatial Encoding, employing a Cartesian and Vector dual-stream strategy, effectively utilizes skeletal information in the form of joint and bone features. The proposed method achieves 92.1% Top-1 accuracy with real-time inference on the benchmark, outperforming the baseline by +5.9% accuracy with only 10% FLOPs. Cross-dataset experiments confirm the superiority of our method in robustness.

Jingmei Yang, Mahtab Talaei, Britta Lassmann, Nahid Bhadelia, Ioannis Ch. Paschalidis

We introduce PandemIQ Llama, a domain-adapted large language model (LLM) designed specifically for pandemic intelligence applications. Building on the pre-trained Llama-3.1-8B model, we conducted continuous training using our curated Pandemic Corpus. This dataset was assembled from authoritative public health sources, scientific literature, and specialized knowledge repositories, comprising 508,924 documents totaling 5.8 billion tokens, which is the largest pandemic domain specific data cohort for LLM training. Benefited from our curated large data cohorts and through continuous training leveraging extensive computational resources, the developed PandemIQ Llama model can extract critical domain knowledge on pandemic, which is typically underrepresented in general-purpose language models, To evaluate its performance, we conducted comprehensive comparison of PandemIQ Llama with both prompt-engineered and task-specific fine-tuned baseline models using two tasks: the Biomedical Alert News Question Answering task (1,508 reports with 30 expert-generated questions each) and the Disease Event Type Classification benchmark (4,500 news snippets across eight disease categories). PandemIQ Llama demonstrated substantial improvements over strong baseline models, achieving performance gains ranging from 3.8% to 10.97%. These results suggest that PandemIQ Llama could significantly enhance public health surveillance and analysis capabilities. In addition, our result also suggests that the LLMs can perform better with continuous training than fine-tuning on domain specific tasks. Social Impact: The BEACON platform, powered by our model, launched and now serves over 100 government and multilateral public health organizations and users across 154 countries. Analytics from the platform is being integrated into the Epidemic Intelligence from Open Sources system run by the World Health Organization. This integration will provide public health decision-makers with a powerful LLM-based tool for pandemic surveillance.

Ziqi Xu, Yi Liu, Yuekang Li, Ling Shi, Kailong Wang, Yongxin Zhao

People who stutter (PWS) face systemic exclusion in today’s voice-driven society, where access to voice assistants, authentication systems, and remote work tools increasingly depends on fluent speech. Current automatic speech recognition (ASR) systems, trained predominantly on fluent speech, fail to serve millions of PWS worldwide. We present STEAMROLLER, a real time system that transforms stuttered speech into fluent output through a novel multi-stage, multi-agent AI pipeline. Our approach addresses three critical technical challenges: (1) the difficulty of direct speech to speech conversion for disfluent input, (2) semantic distortions introduced during ASR transcription of stuttered speech, and (3) latency constraints for real time communication. STEAMROLLER employs a three stage architecture comprising ASR transcription, multi-agent text repair, and speech synthesis, where our core innovation lies in a collaborative multi-agent framework that iteratively refines transcripts while preserving semantic intent. Experiments on the FluencyBank dataset and a user study demonstrates clear word error rate (WER) reduction and strong user satisfaction. Beyond immediate accessibility benefits, fine tuning ASR on STEAMROLLER repaired speech further yields additional WER improvements, creating a pathway toward inclusive AI ecosystems.

Zhengsen Xu, Sibo Cheng, Lanying Wang, Hongjie He, Wentao Sun, Jonathan Li, Lincoln Linlin Xu

Wildfire risk prediction remains a critical yet challenging task due to the complex interactions among fuel conditions, meteorology, topography, and human activity. Despite growing interest in data-driven approaches, publicly available benchmark datasets that support long-term temporal modeling, large-scale spatial coverage, and multimodal drivers remain scarce. To address this gap, we present a 25-year, daily-resolution wildfire dataset covering 240 million hectares across British Columbia and surrounding regions. The dataset includes 38 covariates, encompassing active fire detections, weather variables, fuel conditions, terrain features, and anthropogenic factors. Using this benchmark, we evaluate a diverse set of time-series forecasting models, including CNN-based, linear-based, Transformer-based, and Mamba-based architectures. We also investigate effectiveness of position embedding and the relative importance of different fire-driving factors.

Wenhao Xu, Akshatha Arodi, Jian-Yun Nie, Arsène Fansi Tchango

Modern slavery affects millions of people worldwide, and regulatory frameworks such as Modern Slavery Acts now require companies to publish detailed disclosures. However, these statements are often vague and inconsistent, making manual review time-consuming and difficult to scale. While NLP offers a promising path forward, high-stakes compliance tasks require more than accurate classification: they demand transparent, rule-aligned outputs that legal experts can verify. Existing applications of large language models (LLMs) often reduce complex regulatory assessments to binary decisions, lacking the necessary structure for robust legal scrutiny. We argue that compliance verification is fundamentally a rule-matching problem: it requires evaluating whether textual statements adhere to well-defined regulatory rules. To this end, we propose a novel framework that harnesses AI for rule-level compliance verification while preserving expert oversight. At its core is the Compliance Alignment Judge (CA-Judge), which evaluates model-generated justifications based on their fidelity to statutory requirements. Using this feedback, we train the Compliance Alignment LLM (CALLM), a model that produces rule-consistent, human-verifiable outputs. CALLM improves predictive performance and generates outputs that are both transparent and legally grounded, offering a more verifiable and actionable solution for real-world compliance analysis.

Yixuan Xie, Yang He, Xiaoyu Yang, Xu Gai, Pan Hui

Integrating Large Language Models (LLMs) into judicial decision-making demands rigorous safety examination against non-legal influences. This paper presents a novel stress test where we evaluate LLM-generated labor dispute outcomes by introducing social media sentiment as an external pressure, critically comparing them against 10,000 real-world court judgments from China Judgments Online (CJOL). Our findings reveal significant LLM safety vulnerabilities: models exhibit inherent deviations from real rulings, and public opinion substantially amplifies these discrepancies, leading to unstable and often inflated compensation predictions. Furthermore, these safety risks are compounded across low-skilled occupational categories and emotionally charged topics. This study uncovers critical threats to judicial integrity and public trust, underscoring the urgent need for robust safeguards against non-legal influences in AI legal systems.

Yin Wu, Zhengxuan Zhang, Jiayu Chen, Chang Xu, Yuyu Luo, Nan Tang, Hui Xiong

Oracle Bone Script, East Asia's earliest mature writing system from over 3,500 years ago, encodes ancient cognition through visual metaphors, yet remains largely undeciphered and inaccessible, severing modern society from its cultural roots. Traditional AI methods, while accurate in classification, treat glyphs as opaque data, neglecting their pictographic essence and failing to foster public understanding—exacerbating a heritage crisis amid linguistic evolution. We pioneer a paradigm shift toward AI-driven cultural democratization, introducing OracleVis, the first human-validated multimodal dataset of glyph-image-explanation triplets, curated through expert collaborations to overcome data scarcity, bias, and incompleteness in archaeological sources. Building on this, OBS-VM, an explainability-centric multimodal large language model fine-tuned on Qwen2-VL-7B, models pictographic reasoning by balancing semantic fidelity with interpretive transparency, transforming black-box predictions into cognition-aligned narratives. Rigorous evaluations, including benchmarks and a user study with 24 non-experts, reveal our system's superiority: it outperforms GPT-4o in pictographic rationality (3.79 vs. 3.58 in human evaluation) and achieves a 35.3% relative improvement in recognition accuracy, while interactive learning boosts knowledge gains (+5.5 vs. +1.7), interest (+1.9 vs. +0.4), and confidence (+2.0 vs. +0.3) over static methods. This work illuminates AI's potential to bridge ancient wisdom and contemporary audiences, redefining heritage preservation as an inclusive, socially impactful endeavor that turns cultural alienation into enlightened engagement.

Jinge Wu, Zhaolong Wu, Ruizhe Li, Tong Chen, Abul Hasan, Yunsoo Kim, Jason Pui-Yin Cheung, Teng Zhang, Honghan Wu

The increasing complexity and workload of clinical radiology leads to inevitable oversights and mistakes in their use as diagnostic tools, causing delayed treatments and sometimes life-threatening harm to patients. While large language models (LLMs) have shown remarkable progress in many tasks, their utilities in detecting and correcting errors in radiology reporting are limited. This paper proposes a novel dual-knowledge infusion framework that enhances LLMs' capability for radiology report proofreading through systematic integration of medical expertise. Specifically, the knowledge infusion combines medical knowledge graph distillation (MKGD) with external knowledge retrieval (EXKR), enabling an effective automated approach in tackling mistakes in radiology reporting. By decomposing the complex proofreading task into three specialized stages of detection, localization, and correction, our method mirrors the systematic review process employed by expert radiologists, ensuring both precision and clinical interpretability. To perform a robust, clinically relevant evaluation, a comprehensive benchmark is also proposed using real-world radiology reports with real-world error patterns, including speech recognition confusions, terminology ambiguities, and template-related inconsistencies. Extensive evaluations across multiple LLM architectures demonstrate substantial improvements of our approach: up to 31.56% increase in error detection accuracy and 37.4% reduction in processing time. Human evaluation by radiologists confirms superior clinical relevance and factual consistency compared to existing approaches.

Penghui Wen, Mengwei He, Patrick Filippi, Na Zhao, Feng Zhang, Thomas Francis Bishop, Zhiyong Wang, Kun Hu

Accurate short-term precipitation forecasting is critical for weather-sensitive decision-making in agriculture, transportation, and disaster response. Existing deep learning approaches often struggle to balance global structural consistency with local detail preservation, especially under complex meteorological conditions. We propose DuoCast, a dual-diffusion framework that decomposes precipitation forecasting into low- and high-frequency components modeled in orthogonal latent subspaces. We theoretically prove that this frequency decomposition reduces prediction error compared to conventional single branch U-Net diffusion models. In DuoCast, the low-frequency model captures large-scale trends via convolutional encoders conditioned on weather front dynamics, while the high-frequency model refines fine-scale variability using a self-attention-based architecture. Experiments on four benchmark radar datasets show that DuoCast consistently outperforms state-of-the-art baselines, achieving superior accuracy in both spatial detail and temporal evolution.

Keshara Weerasinghe, Xueren Ge, Tessa Heick, Lahiru Nuwan Wijayasingha, Anthony Cortez, Abhishek Satpathy, John Stankovic, Homa Alemzadeh

Emergency Medical Services (EMS) are critical to patient survival in emergencies, but first responders often face intense cognitive demands in high-stakes situations. AI cognitive assistants, acting as virtual partners, have the potential to ease this burden by supporting real-time data collection and decision making. In pursuit of this vision, we introduce EgoEMS, the first end-to-end, high-fidelity, multimodal, multiperson dataset capturing over 20 hours of realistic, procedural EMS activities from an egocentric view in 233 simulated emergency scenarios performed by 62 participants, including 46 EMS professionals. Developed in collaboration with EMS experts and aligned with national standards, EgoEMS is captured using an open-source, low-cost, and replicable data collection system and is annotated with keysteps, timestamped audio transcripts with speaker diarization, action quality metrics, and bounding boxes with segmentation masks. Emphasizing realism, the dataset includes responder-patient interactions reflecting real-world emergency dynamics. We also present a suite of benchmarks for real-time multimodal keystep recognition and action quality estimation, essential for developing AI support tools for EMS. We hope EgoEMS inspires the research community to push the boundaries of intelligent EMS systems and ultimately contribute to improved patient outcomes.

Zichong Wang, Jie Yang, Jun Zhuang, Puqing Jiang, Mingzhe Chen, Ye Hu, Wenbin Zhang

Graph neural networks (GNNs) excel at modeling graph-structured data but often inherit and amplify biases, leading to substantial efforts in developing fair GNNs. However, most existing approaches assume full access to sensitive attribute information, which is often impractical in real-world scenarios due to privacy concerns or risks of discrimination. To address this limitation, this paper focuses on graph fairness with limited sensitive attribute information, ensuring applicability to real-world contexts where current methods fall short. Specifically, we introduce an innovative fairness optimization strategy, propose a novel framework named FGLISA, and provide a theoretical perspective linking limited sensitive attribute information access to fairness objectives, thus enabling fair graph learning in real-world applications with limited sensitive attribute information. Experiments on diverse real-world datasets and tasks validate the effectiveness of our approach in achieving both fairness and predictive performance.

Zeqiang Wang, Rebecca Oldroyd, Yuqi Wang, Jiageng Wu, Jie Yang, Wei Wang, Nishanth R. Sastry, Jon Johnson, Suparna De

Automated classification of complex social survey questionnaires is crucial for large-scale social science research but faces significant reliability challenges due to intricate hierarchical label structures, severe class imbalance, semantic ambiguity, and incomplete data coverage. Conventional classification methods often struggle with these combined complexities, yielding results that lack trustworthiness. We introduce HOCM, a framework designed for trustworthy classification in complex, real-world taxonomies. It features two synergistic components: (1) memory-enhanced contrastive learning, tailored to learn robust representations from noisy, imbalanced data by leveraging quality-aware category memory banks; and (2) hierarchical uncertainty calibration, which enforces taxonomic consistency while providing reliable confidence estimates and identifying inputs falling outside well-represented known categories. Our evaluation on a large-scale, real-world social survey dataset—a challenging exemplar of our target problem class—demonstrates that HOCM maintains strong accuracy on known classes while effectively identifying uncertain cases, significantly boosting accuracy on confident predictions. Furthermore, it adeptly detects low-resource/unknown categories. HOCM provides a more reliable automated classification tool, enabling efficient expert review and enhancing the trustworthiness of analysis in domains with complex, hierarchical data.

Yingying Wang, Qin Ni, Haoxin Xu, Jiaqi Yin, Tingjiang Wei

Artificial intelligence is playing an increasingly important role in supporting decision-making, particularly in educational contexts, where it serves as a critical tool to assist teacher judgment and optimize instructional decisions. However, limited research has examined how different AI-assisted decision-making paradigms influence the Performance of human-AI collaboration, as well as the underlying psychological mechanisms and causal pathways. Therefore, this study investigated 59 pre-service teachers to examine how AI-assisted decision-making paradigms and human-AI consistency influenced their psychological states and task performance. Specifically, this study employed a two-factor mixed experimental design, with the AI-assisted decision-making paradigms as the between-subjects factor and human-AI consistency as the within-subjects factor. Data were analyzed using the Bayesian cumulative link mixed model and structural equation modeling. The results reveal that AI-assisted decision-making paradigms do not have a significant direct effect on task performance. However, when the moderating role of human-AI decision consistency is taken into account, the effect of AI-assisted decision-making paradigms on task performance can exert its influence indirectly through a sequential psychological pathway involving users’ confidence and their trust in the AI. Consistency between human and AI decisions not only significantly enhances users’ trust in AI, confidence, and task performance, but the proportion of consistent decisions also significantly moderates the impact of AI-assisted decision-making paradigms on users’ confidence levels. Notably, our findings indicate that users maintain a moderately level of trust in AI even when their decisions diverge from those of AI. In summary, this study highlights the mediating mechanism by which AI-assisted decision-making paradigms influence task performance through psychological states and identifies the moderating role of human-AI consistency in this pathway. These findings advance the theoretical understanding of human-AI interaction models in educational contexts and offer mechanistic insights to guide the optimization of instructional AI systems.

Peijian Wang, Thu Nguyen

For the past two decades, sustainability and carbon reduction have emerged as critical factors for data center (DC) design and operation. A set of advances has been encapsulated in green DCs with the onsite generation of renewable energy and efficient cooling systems. In this paper, we study how to apply Deep Reinforcement Learning (DRL) to optimize green DC operation. Green DC management is typically an infinite-horizon problem with exogenous stochastic input processes. We propose EA, a framework that applies DRL to the typical infinite-horizon problem without discounting. EA approximates the infinite-horizon problem with a finite-horizon one. In this approach, it is important to avoid actions optimized for the end of the finite-horizon problem but inappropriate for the true infinite-horizon one. EA addresses this challenge by combining a stationary policy with the fact that green DC management has repeating patterns (e.g., daily temperature, solar energy generation, and workload). We apply EA to the management of a green DC with onsite solar energy generation and a hybrid cooling system that includes ``free'' cooling. Evaluation results show that EA successfully learns important principles such as delaying deferrable jobs to solar-rich times and gracefully maintaining inside temperature. Further, EA outperforms three state-of-the-art DRL algorithms, realizing the greatest benefits on days with high outside temperature and high solar generation. While we evaluate EA in the specific context of a green DC, we believe that EA is a promising approach for more general system management settings.

Han Wang, Deyi Ji, Junyu Lu, Lanyun Zhu, Hailong Zhang, Haiyang Wu, Liqun Liu, Peng Shu, Roy Ka-Wei Lee

Accurate detection of offensive content on social media demands high-quality labeled data; however, such data is often scarce due to the low prevalence of offensive instances and the high cost of manual annotation. To address this low-resource challenge, we propose a self-training framework that leverages abundant unlabeled data through collaborative pseudo-labeling. Starting with a lightweight classifier trained on limited labeled data, our method iteratively assigns pseudo-labels to unlabeled instances with the support of Multi-Agent Vision-Language Models (MA-VLMs). Unlabeled data on which the classifier and MA-VLMs agree are designated as the Agreed-Unknown set, while conflicting samples form the Disagreed-Unknown set. To enhance label reliability, MA-VLMs simulate dual perspectives, moderator and user, capturing both regulatory and subjective viewpoints. The classifier is optimized using a novel Positive-Negative-Unlabeled (PNU) loss, which jointly exploits labeled, Agreed-Unknown, and Disagreed-Unknown data while mitigating pseudo-label noise. Experiments on benchmark datasets demonstrate that our framework substantially outperforms baselines under limited supervision and approaches the performance of large-scale models.