arXiv subjects

Pengfei Wei

Publications and source records attributed to Pengfei Wei.

At least 19 recordsLinked to original sources

arXiv2026

When Models Hear What They Expect: Diagnosing Prosodic Heuristics in Multimodal Sarcasm Detection

Multimodal Large Language Models (MLLMs) process speech and text jointly, yet whether they exploit prosodic cues for pragmatic inference or rely on surface acoustic patterns has received little systematic investigation. We address this through sarcasm detection, evaluating Qwen2.5-Omni and Qwen3-Omni on Mandarin Chinese and English under five modality conditions that decompose the contributions of lexical content, vocal semantics, and prosodic structure. Adding audio systematically inflates false positives without improving true positive detection. Acoustic error diagnosis reveals that model errors cluster on a shared stereotype of expressive prosody, namely elevated pitch and irregular pausing, that diverges from the actual cues marking sarcasm in both languages. Targeted manipulation of only these two dimensions causally confirms the heuristic, inducing false positive rates of up to 60%. Applying the same manipulation template to Gemini~3 Flash Preview without modification replicates the effect, suggesting that the stereotype extends beyond the Qwen Omni family rather than arising from a single model architecture.

cs.CL↗

arXiv2026

GGSS: Geodesic-Gated Spherical Steering for Inference-Time Debiasing of Generative Vision-Language Models

Generative vision-language models (VLMs) are increasingly used in human-centered settings, yet they can produce demographically biased outputs even when images differ only in controlled attributes such as perceived race or gender. However, existing inference-time debiasers were largely designed for static embeddings or CLIP-like models rather than generative VLMs. We propose GGSS---Geodesic-Gated Spherical Steering---a norm-preserving intervention that discovers a counterfactual bias subspace on the unit hypersphere, steers visual tokens along geodesic arcs, and uses an adaptive gate to focus correction on tokens that carry stronger demographic signal. We evaluate four generative VLMs against ten adapted inference-time debiasing baselines and prompt-based mitigation under a single operating-point protocol across categorical, pairwise, and occupation-gender bias tests, while also measuring general visual-language capability. GGSS achieves the lowest average bias on all four models, significant on three of four backbones under paired permutation tests, while preserving MMStar accuracy within +/- 0.6 p.p. of the unsteered baseline. Code is available at https://github.com/dukesun99/GGSS.

cs.CY↗

arXiv2026

OpTI-Mouse: Optimization for Targeted Temporal Interference Stimulation in the Mouse Brain

Temporal Interference (TI) stimulation enables deep brain targeting, yet precise optimization tools for mouse models remain limited. We developed a computational optimization tool integrating mouse head modeling with the optimization algorithm to optimize stimulation strategies for predefined target regions. By balancing target intensity and spatial focality, the optimized strategy significantly outperformed empirical baselines. For the CA3-CA1 target, it achieved a 7-fold intensity increase (10.29 vs. 2.89 V/m) under iso-focality conditions. Conversely, for the Dentate Gyrus, it improved spatial confinement ($r_{0.5}$ reduced from 3.99 to 3.54 mm) while maintaining comparable intensity. Cross-model validation on a standardized Sim4Life phantom further confirmed the framework's robustness. This approach offers a powerful tool for enhancing the precision and reproducibility of preclinical TI stimulation studies.

q-bio.NC↗

arXiv2026

Hybrid Uncertainty Sensitivity Analysis Based on the HSIC for High-Dimensional Responses with Aleatory--Epistemic Separation

Quantifying the influence of hybrid aleatory and epistemic uncertainties on high-dimensional system responses remains a major challenge in global sensitivity analysis (GSA). Existing Hilbert--Schmidt Independence Criterion (HSIC)-based approaches are primarily restricted to single-output settings and lack a rigorous decomposition of heterogeneous uncertainty sources and their interactions. To address this limitation, a novel double-space tensor-product RKHS framework is proposed for sensitivity analysis under hybrid uncertainty. By constructing factorized kernels over both the latent input space and the multidimensional output space, a concurrent double M\"obius inversion is derived to orthogonally decompose the global dependence measure into pure aleatory effects, pure epistemic effects, and their interaction contributions. The resulting dimension-wise sensitivity indices preserve the uncertainty attribution structure across all output dimensions. To satisfy the independence assumptions required by the decomposition, an auxiliary-variable representation based on the inverse probability integral transform is introduced, enabling the treatment of hierarchical uncertainties and Copula-induced correlations within a unified latent space. A fully vectorized single-loop implementation is further developed to avoid the computational burden of nested Monte Carlo simulation. Statistical significance and estimation uncertainty are quantified through permutation testing and Bootstrap confidence intervals. Numerical studies on a modified multi-output Ishigami function and an aerodynamic pressure-field problem demonstrate the accuracy, scalability, and practical applicability of the proposed framework.

stat.ML↗

arXiv2026

Efficient Temporal Modeling for Mobile Sleep Staging via Lightweight Random Attention

Mobile sleep staging serves as a foundational infrastructure for in-home sleep monitoring and closed-loop modulation. But existing sequential models such as RNNs and Transformers are computationally expensive for mobile deployment. In this paper, we propose Random Attention (RA), a lightweight temporal modeling module based on fixed random projections, which replaces learnable sequence modeling with similarity-based aggregation. RA introduces little additional parameters beyond the epoch encoder while enabling effective temporal smoothing. We further provide a theoretical interpretation via the Random Attention Prior Kernel (RAPK), which decomposes RA into a global smoothing term and a feature similarity term, offering an interpretable view of temporal sleep structure. Experiments on Sleep-EDF-20 and Sleep-EDF-78 show that RA consistently improves epoch-wise baselines by 1-3\% in accuracy and F1 score, while achieving competitive performance compared with LSTM, GRU, and Transformer models. RA also demonstrates strong generalization across different backbone encoders and improved robustness over conventional temporal smoothing methods. These results indicate that efficient sleep staging can be achieved through lightweight similarity-based temporal aggregation, making RA suitable for real-time wearable applications.

eess.SP↗

arXiv2026

Return of Frustratingly Easy Unsupervised Video Domain Adaptation

Unsupervised video domain adaptation (UVDA) is a practical but under-explored problem. In this paper, we propose a frustratingly easy UVDA method, called MetaTrans. Specifically, MetaTrans adopts a concise learning objective that contains only two fundamental loss terms. Despite the simplicity of the learning objective, MetaTrans embodies an advanced UVDA idea, that is, handling the spatial and temporal divergence of cross-domain videos separately, through a subtle model architecture design. By implementing a temporal-static subtraction module, MetaTrans effectively removes spatial and temporal divergence. Extensive empirical evaluations, particularly on various cross-domain action recognition tasks, show substantial absolute adaptation performance enhancement and significantly superior relative performance gain compared with state-of-the-art UVDA baselines.

cs.CV↗

arXiv2026

Very Efficient Listwise Multimodal Reranking for Long Documents

Listwise reranking is a key yet computationally expensive component in vision-centric retrieval and multimodal retrieval-augmented generation (M-RAG) over long documents. While recent VLM-based rerankers achieve strong accuracy, their practicality is often limited by long visual-token sequences and multi-step autoregressive decoding. We propose ZipRerank, a highly efficient listwise multimodal reranker that directly addresses both bottlenecks. It reduces input length via a lightweight query-image early interaction mechanism and eliminates autoregressive decoding by scoring all candidates in a single forward pass. To enable effective learning, ZipRerank adopts a two-stage training strategy: (i) listwise pretraining on large-scale text data rendered as images, and (ii) multimodal finetuning with VLM-teacher-distilled soft-ranking supervision. Extensive experiments on the MMDocIR benchmark show that ZipRerank matches or surpasses state-of-the-art multimodal rerankers while reducing LLM inference latency by up to an order of magnitude, making it well-suited for latency-sensitive real-world systems. The code is available at https://github.com/dukesun99/ZipRerank.

cs.IR↗

arXiv2026

Rethinking Random Transformers as Adaptive Sequence Smoothers for Sleep Staging

Automatic sleep staging commonly adopts Transformers under the assumption that they learn complex long-range dependencies. We challenge this view by revealing a neglected property of sleep sequences: strong local temporal continuity. We show that a randomly initialized Transformer, without any training, substantially improves sleep staging performance and consistently outperforms heuristic smoothing. We formalize this effect via a Random Attention Prior Kernel (RAPK), showing that random self-attention acts as an adaptive smoother by balancing global averaging and content-based similarity while preserving stage transitions. Using two metrics, the Local Smoothness Influence Index (LSII) and the Weighted Transition Entropy (WTE), we provide evidence that most performance gains in Transformer-based sleep staging arise from architectural inductive bias rather than parameter learning. Our results suggest that sleep staging can be effectively addressed with structure-driven smoothing mechanisms rather than complex dependency modeling, enabling more efficient and edge-deployable healthcare systems for large-scale physiological monitoring.

cs.LG↗

arXiv2026

Sparse Autoencoders as Plug-and-Play Firewalls for Adversarial Attack Detection in VLMs

Vision-language models (VLMs) have advanced rapidly and are increasingly deployed in real-world applications, especially with the rise of agent-based systems. However, their safety has received relatively limited attention. Even the latest proprietary and open-weight VLMs remain highly vulnerable to adversarial attacks, leaving downstream applications exposed to significant risks. In this work, we propose a novel and lightweight adversarial attack detection framework based on sparse autoencoders (SAEs), termed SAEgis. By inserting an SAE module into a pretrained VLM and training it with standard reconstruction objectives, we find that the learned sparse latent features naturally capture attack-relevant signals. These features enable reliable classification of whether an input image has been adversarially perturbed, even for previously unseen samples. Extensive experiments show that SAEgis achieves strong performance across in-domain, cross-domain, and cross-attack settings, with particularly large improvements in cross-domain generalization compared to existing baselines. In addition, combining signals from multiple layers further improves robustness and stability. To the best of our knowledge, this is the first work to explore SAE as a plug-and-play mechanism for adversarial attack detection in VLMs. Our method requires no additional adversarial training, introduces minimal overhead, and provides a practical approach for improving the safety of real-world VLM systems.

cs.CV↗

arXiv2026

Corpus2Skill: Distilling Enterprise Knowledge into Navigable Agent Skills for QA and RAG

Retrieval-Augmented Generation (RAG) grounds LLM responses in external evidence but treats the model as a passive consumer of search results, with no view of how the corpus is organized or what it has not yet seen. We present Corpus2Skill, a system-level retrieval architecture for bounded, structurally coherent corpora such as enterprise knowledge bases: an offline compiler distills the corpus into a hierarchical skill directory, and at serve time an LLM agent navigates it, drilling from a bird's-eye view through progressively finer summaries down to documents and backtracking when a branch is unproductive. On an enterprise customer-support benchmark, Corpus2Skill improves both answer quality and grounding over single-shot dense, hybrid, hierarchical-retrieval, and agentic RAG baselines at a moderate cost tradeoff, and the lead persists under encoder-matched controls and paired significance tests. An eleven-dataset study shows that corpus navigation is not a universal replacement for retrieval: it significantly wins on five datasets, ties on three, and loses on three. It helps on single-domain corpora with a recoverable topical taxonomy, but flat retrieval remains preferable on open-domain factoid pools or homogeneous-tabular corpora that defeat top-level clustering. We characterize this scope distinction as a design guideline for knowledge-grounded systems. Code is available at https://github.com/dukesun99/Corpus2Skill.

cs.IR↗

arXiv2026

A Rapid Instrument Exchange System for Humanoid Robots in Minimally Invasive Surgery

Humanoid robot technologies have demonstrated immense potential for minimally invasive surgery (MIS). Unlike dedicated multi-arm surgical platforms, the inherent dual-arm configuration of humanoid robots necessitates an efficient instrument exchange capability to perform complex procedures, mimicking the natural workflow where surgeons manually switch instruments. To address this, this paper proposes an immersive teleoperated rapid instrument exchange system. The system utilizes a low-latency mechanism based on single-axis compliant docking and environmental constraint release. Integrated with real-time first-person view (FPV) perception via a head-mounted display (HMD), this framework significantly reduces operational complexity and cognitive load during the docking process. Comparative evaluations between experts and novices demonstrate high operational robustness and a rapidly converging learning curve; novice performance in instrument attachment and detachment improved substantially after brief training. While long-distance spatial alignment still presents challenges in time cost and collaborative stability, this study successfully validates the technical feasibility of humanoid robots executing stable instrument exchanges within constrained clinical environments.

cs.RO↗

arXiv2026

OrchMAS: Orchestrated Reasoning with Multi Collaborative Heterogeneous Scientific Expert Structured Agents

Multi-agent large language model frameworks are promising for complex multi step reasoning, yet existing systems remain weak for scientific and knowledge intensive domains due to static prompts and agent roles, rigid workflows, and homogeneous model reliance, leading to poor domain adaptation, limited reasoning flexibility, and high latency on heterogeneous or long-horizon scientific tasks. They also struggle to revise earlier decisions when intermediate reasoning diverges, reducing reliability in structured and calculation heavy settings. To address these limitations, we propose a scientific domain oriented interactive two tier multi model orchestration framework. A dedicated orchestration model analyzes each task, dynamically constructs a domain aware reasoning pipeline, and instantiates specialized expert agents with tailored prompts, while an execution model performs each step under generated role and instruction specifications. The orchestrator iteratively updates the pipeline based on intermediate feedback, enabling dynamic replanning, role reallocation, and prompt refinement across multi turn interactions, strengthening robustness and specialization for scientific reasoning through structured heterogeneous model collaboration. The framework is model agnostic and supports heterogeneous LLM integration with different capacities or costs, enabling flexible performance efficiency trade offs in practical scientific deployments. Experiments show consistent improvements over existing multi agent systems and strong baselines across diverse reasoning and scientific style benchmarks.

cs.AI↗

arXiv2025

Sleep Modulation: The Challenge of Transitioning from Open Loop to Closed Loop

Sleep disorders have emerged as a critical global health issue, highlighting the urgent need for effective and widely accessible intervention technologies. Non-invasive brain stimulation has garnered attention as it enables direct or indirect modulation of neural activity, thereby promoting sleep enhancement in a safe and unobtrusive manner. This class of approaches is collectively referred to as sleep modulation. To date, the majority of sleep modulation research relies on open-loop paradigms with empirically determined parameters, while achieving individual adaptation and modulation accuracy remains a distant objective. The paradigm-specific constraints inherent to open-loop designs represent a major obstacle to clinical translation and large-scale deployment in home environments. In this paper, we delineate fundamental paradigms of sleep modulation, critically examine the intrinsic limitations of open-loop approaches, and formally conceptualize sleep closed-loop modulation. We further provide a comprehensive synthesis of prior studies involving five commonly employed modulation techniques, evaluating their potential integration within a closed-loop framework. Finally, we identify three primary challenges in constructing an effective sleep closed-loop modulation system: sensor solution selection, monitoring model design, and modulation strategy design, while also proposing potential solutions. Collectively, this work aims to advance the paradigm shift of sleep modulation from open-loop toward closed-loop systems.

cs.HC↗

arXiv2025

Bayesian Active Learning for Bayesian Model Updating: the Art of Acquisition Functions and Beyond

Estimating posteriors and the associated model evidences, with desired accuracy and affordable computational cost, is a core issue of Bayesian model updating, and can be of great challenge given expensive-to-evaluate models and posteriors with complex features such as multi-modalities of unequal importance, nonlinear dependencies and high sharpness. Bayesian Quadrature (BQ) equipped with active learning has emerged as a competitive framework for tackling this challenge, as it provides flexible balance between computational cost and accuracy. The performance of a BQ scheme is fundamentally dictated by the acquisition function as it exclusively governs the active generation of integration points. After reexamining one of the most advanced acquisition function from a prospective inference perspective and reformulating the quadrature rules for prediction, four new acquisition functions, inspired by distinct intuitions on expected rewards, are primarily developed, all of which are accompanied by elegant interpretations and highly efficient numerical estimators. Mathematically, these four acquisition functions measure, respectively, the prediction uncertainty of posterior, the contribution to prediction uncertainty of evidence, as well as the expected reduction of prediction uncertainties concerning posterior and evidence, and thus provide flexibility for highly effective design of integration points. These acquisition functions are further extended to the transitional BQ scheme, along with several specific refinements, to tackle the above-mentioned challenges with high efficiency and robustness. Effectiveness of the developments is ultimately demonstrated with extensive benchmark studies and application to an engineering example.

stat.CO↗

arXiv2025

Personalized Transcranial Electrical Stimulation: A Review of Computational Modeling and Optimization

Objective. Personalized transcranial electrical stimulation (tES) has gained growing attention due to the substantial inter-individual variability in brain anatomy and physiology. While previous reviews have discussed the physiological mechanisms and clinical applications of tES, there remains a critical gap in up-to-date syntheses focused on the computational modeling frameworks that enable individualized stimulation optimization. Approach. This review presents a comprehensive overview of recent advances in computational techniques supporting personalized tES. We systematically examine developments in forward modeling for simulating individualized electric fields, as well as inverse modeling approaches for optimizing stimulation parameters. We critically evaluate progress in head modeling pipelines, optimization algorithms, and the integration of multimodal brain data. Main results. Recent advances have substantially accelerated the construction of subject-specific head conductor models and expanded the landscape of optimization methods, including multi-objective optimization and brain network-informed optimization. These advances allow for dynamic and individualized stimulation planning, moving beyond empirical trial-and-error approaches.Significance. By integrating the latest developments in computational modeling for personalized tES, this review highlights current challenges, emerging opportunities, and future directions for achieving precision neuromodulation in both research and clinical contexts.

q-bio.NC↗

arXiv2025

DreamActor-H1: High-Fidelity Human-Product Demonstration Video Generation via Motion-designed Diffusion Transformers

In e-commerce and digital marketing, generating high-fidelity human-product demonstration videos is important for effective product presentation. However, most existing frameworks either fail to preserve the identities of both humans and products or lack an understanding of human-product spatial relationships, leading to unrealistic representations and unnatural interactions. To address these challenges, we propose a Diffusion Transformer (DiT)-based framework. Our method simultaneously preserves human identities and product-specific details, such as logos and textures, by injecting paired human-product reference information and utilizing an additional masked cross-attention mechanism. We employ a 3D body mesh template and product bounding boxes to provide precise motion guidance, enabling intuitive alignment of hand gestures with product placements. Additionally, structured text encoding is used to incorporate category-level semantics, enhancing 3D consistency during small rotational changes across frames. Trained on a hybrid dataset with extensive data augmentation strategies, our approach outperforms state-of-the-art techniques in maintaining the identity integrity of both humans and products and generating realistic demonstration motions. Project page: https://lizhenwangt.github.io/DreamActor-H1/.

cs.CV↗

arXiv2025

PoseTraj: Pose-Aware Trajectory Control in Video Diffusion

Recent advancements in trajectory-guided video generation have achieved notable progress. However, existing models still face challenges in generating object motions with potentially changing 6D poses under wide-range rotations, due to limited 3D understanding. To address this problem, we introduce PoseTraj, a pose-aware video dragging model for generating 3D-aligned motion from 2D trajectories. Our method adopts a novel two-stage pose-aware pretraining framework, improving 3D understanding across diverse trajectories. Specifically, we propose a large-scale synthetic dataset PoseTraj-10K, containing 10k videos of objects following rotational trajectories, and enhance the model perception of object pose changes by incorporating 3D bounding boxes as intermediate supervision signals. Following this, we fine-tune the trajectory-controlling module on real-world videos, applying an additional camera-disentanglement module to further refine motion accuracy. Experiments on various benchmark datasets demonstrate that our method not only excels in 3D pose-aligned dragging for rotational trajectories but also outperforms existing baselines in trajectory accuracy and video quality.

cs.CV↗

arXiv2024

Sampling and active learning methods for network reliability estimation using K-terminal spanning tree

Network reliability analysis remains a challenge due to the increasing size and complexity of networks. This paper presents a novel sampling method and an active learning method for efficient and accurate network reliability estimation under node failure and edge failure scenarios. The proposed sampling method adopts Monte Carlo technique to sample component lifetimes and the K-terminal spanning tree algorithm to accelerate structure function computation. Unlike existing methods that compute only one structure function value per sample, our method generates multiple component state vectors and corresponding structure function values from each sample. Network reliability is estimated based on survival signatures derived from these values. A transformation technique extends this method to handle both node failure and edge failure. To enhance efficiency of proposed sampling method and achieve adaptability to network topology changes, we introduce an active learning method utilizing a random forest (RF) classifier. This classifier directly predicts structure function values, integrates network behaviors across diverse topologies, and undergoes iterative refinement to enhance predictive accuracy. Importantly, the trained RF classifier can directly predict reliability for variant networks, a capability beyond the sampling method alone. Through investigating several network examples and two practical applications, the effectiveness of both proposed methods is demonstrated.

cs.LG↗