# Research Papers — 2026-03 · bundle page 33 of 243 > Full-text markdown of papers #1601–1650 of 12123 published in 2026-03, newest first. AI-readable mirrors of the full 12123-paper 2026-03 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-03/page/33 [← Previous bundle](https://aipod.dev/papers/archive/2026-03/page/32.md) [Next bundle →](https://aipod.dev/papers/archive/2026-03/page/34.md) ### Dual-Stage Invariant Continual Learning under Extreme Visual Sparsity Authors: Rangya Zhang, Jiaping Xiao, Lu Bai, Yuhang Zhang, Mir Feroskhan Published: 2026-03-27 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2603.26190v1 Continual learning seeks to maintain stable adaptation under non-stationary environments, yet this problem becomes particularly challenging in object detection, where most existing methods implicitly assume relatively balanced visual conditions. In extreme-sparsity regimes, such as those observed in space-based resident space object (RSO) detection scenarios, foreground signals are overwhelmingly dominated by background observations. Under such conditions, we analytically demonstrate that background-driven gradients destabilize the feature backbone during sequential domain shifts, causing progressive representation drift. This exposes a structural limitation of continual learning approaches relying solely on output-level distillation, as they fail to preserve intermediate representation stability. To address this, we propose a dual-stage invariant continual learning framework via joint distillation, enforcing structural and semantic consistency on both backbone representations and detection predictions, respectively, thereby suppressing error propagation at its source while maintaining adaptability. Furthermore, to regulate gradient statistics under severe imbalance, we introduce a sparsity-aware data conditioning strategy combining patch-based sampling and distribution-aware augmentation. Experiments on a high-resolution space-based RSO detection dataset show consistent improvement over established continual object detection methods, achieving an absolute gain of +4.0 mAP under sequential domain shifts. --- ### OSA: Echocardiography Video Segmentation via Orthogonalized State Update and Anatomical Prior-aware Feature Enhancement Authors: Rui Wang, Huisi Wu, Jing Qin Published: 2026-03-27 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.26188v1 Accurate and temporally consistent segmentation of the left ventricle from echocardiography videos is essential for estimating the ejection fraction and assessing cardiac function. However, modeling spatiotemporal dynamics remains difficult due to severe speckle noise and rapid non-rigid deformations. Existing linear recurrent models offer efficient in-context associative recall for temporal tracking, but rely on unconstrained state updates, which cause progressive singular value decay in the state matrix, a phenomenon known as rank collapse, resulting in anatomical details being overwhelmed by noise. To address this, we propose OSA, a framework that constrains the state evolution on the Stiefel manifold. We introduce the Orthogonalized State Update (OSU) mechanism, which formulates the memory evolution as Euclidean projected gradient descent on the Stiefel manifold to prevent rank collapse and maintain stable temporal transitions. Furthermore, an Anatomical Prior-aware Feature Enhancement module explicitly separates anatomical structures from speckle noise through a physics-driven process, providing the temporal tracker with noise-resilient structural cues. Comprehensive experiments on the CAMUS and EchoNet-Dynamic datasets show that OSA achieves state-of-the-art segmentation accuracy and temporal stability, while maintaining real-time inference efficiency for clinical deployment. Codes are available at https://github.com/wangrui2025/OSA. --- ### Progressive Learning with Anatomical Priors for Reliable Left Atrial Scar Segmentation from Late Gadolinium Enhancement MRI Authors: Jing Zhang, Bastien Bergere, Emilie Bollache, Jonas Leite, Mikaël Laredo, Alban Redheuil, Nadjia Kachenoura Published: 2026-03-27 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2603.26186v1 Cardiac MRI late gadolinium enhancement (LGE) enables non-invasive identification of left atrial (LA) scar, whose spatial distribution is strongly associated with atrial fibrillation (AF) severity and recurrence. However, automatic LA scar segmentation remains challenging due to low contrast, annotation variability, and the lack of anatomical constraints, often leading to non-reliable predictions. Accordingly, our aim was to propose a progressive learning strategy to segment LA scar from LGE images inspired from a clinical workflow. A 3-stage framework based on SwinUNETR was implemented, comprising: 1) a first LA cavity pre-learning model, 2) dual-task model which further learns spatial relationship between LA geometry and scar patterns, and 3) fine-tuning on precise segmentation of the scar. Furthermore, we introduced an anatomy-aware spatially weighted loss that incorporates prior clinical knowledge by constraining scar predictions to anatomically plausible LA wall regions while mitigating annotation bias. Our preliminary results obtained on validation LGE volumes from LASCARQS public dataset after 5-fold cross validation, LA segmentation had Dice score of 0.94, LA scar segmentation achieved Dice score of 0.50, Hausdorff Distance of 11.84 mm, Average Surface Distance of 1.80 mm, outperforming only a one-stage scar segmentation with 0.49, 13.02 mm, 1.96 mm, repectively. By explicitly embedding clinical anatomical priors and diagnostic reasoning into deep learning, the proposed approach improved the accuracy and reliability of LA scar segmentation from LGE, revealing the importance of clinically informed model design. --- ### DUGAE: Unified Geometry and Attribute Enhancement via Spatiotemporal Correlations for G-PCC Compressed Dynamic Point Clouds Authors: Pan Zhao, Hui Yuan, Chang Sun, Chongzhen Tian, Raouf Hamzaoui, Sam Kwong Published: 2026-03-27 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.26183v1 Existing post-decoding quality enhancement methods for point clouds are designed for static data and typically process each frame independently. As a result, they cannot effectively exploit the spatiotemporal correlations present in point cloud sequences.We propose a unified geometry and attribute enhancement framework (DUGAE) for G-PCC compressed dynamic point clouds that explicitly exploits inter-frame spatiotemporal correlations in both geometry and attributes. First, a dynamic geometry enhancement network (DGE-Net) based on sparse convolution (SPConv) and feature-domain geometry motion compensation (GMC) aligns and aggregates spatiotemporal information. Then, a detail-aware k-nearest neighbors (DA-KNN) recoloring module maps the original attributes onto the enhanced geometry at the encoder side, improving mapping completeness and preserving attribute details. Finally, a dynamic attribute enhancement network (DAE-Net) with dedicated temporal feature extraction and feature-domain attribute motion compensation (AMC) refines attributes by modeling complex spatiotemporal correlations. On seven dynamic point clouds from the 8iVFB v2, Owlii, and MVUB datasets, DUGAE significantly enhanced the performance of the latest G-PCC geometry-based solid content test model (GeS-TM v10). For geometry (D1), it achieved an average BD-PSNR gain of 11.03 dB and a 93.95% BD-bitrate reduction. For the luma component, it achieved a 4.23 dB BD-PSNR gain with a 66.61% BD-bitrate reduction. DUGAE also improved perceptual quality (as measured by PCQM) and outperformed V-PCC. Our source code will be released on GitHub at: https://github.com/yuanhui0325/DUGAE --- ### ClinicalAgents: Multi-Agent Orchestration for Clinical Decision Making with Dual-Memory Authors: Zhuohan Ge, Haoyang Li, Yubo Wang, Nicole Hu, Chen Jason Zhang, Qing Li Published: 2026-03-27 Categories: cs.CL arXiv: https://arxiv.org/abs/2603.26182v2 While Large Language Models (LLMs) have demonstrated potential in healthcare, they often struggle with the complex, non-linear reasoning required for accurate clinical diagnosis. Existing methods typically rely on static, linear mappings from symptoms to diagnoses, failing to capture the iterative, hypothesis-driven reasoning inherent in human clinicians. To bridge this gap, we introduce ClinicalAgents, a novel multi-agent framework designed to simulate the cognitive workflow of expert clinicians. Unlike rigid sequential chains, ClinicalAgents employs a dynamic orchestration mechanism modeled as a Monte Carlo Tree Search (MCTS) process. This allows an orchestrator to iteratively generate hypotheses, actively verify evidence, and trigger backtracking when critical information is missing. The foundation of this framework is a Dual-Memory architecture: a mutable working memory that maintains the evolving patient state for context-aware reasoning, and a static experience memory that retrieves clinical guidelines and historical cases via an active feedback loop. Extensive experiments demonstrate that ClinicalAgents achieves the best performance among evaluated baselines, significantly enhancing both diagnostic accuracy and explainability compared to strong single-agent and multi-agent baselines. Our code is released at https://github.com/ZhuohanGe/ClinicalAgents-Code. --- ### GLINT: Modeling Scene-Scale Transparency via Gaussian Radiance Transport Authors: Youngju Na, Jaeseong Yun, Soohyun Ryu, Hyunsu Kim, Sung-Eui Yoon, Suyong Yeon Published: 2026-03-27 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.26181v1 While 3D Gaussian splatting has emerged as a powerful paradigm, it fundamentally fails to model transparency such as glass panels. The core challenge lies in decoupling the intertwined radiance contributions from transparent interfaces and the transmitted geometry observed through the glass. We present GLINT, a framework that models scene-scale transparency through explicit decomposed Gaussian representation. GLINT reconstructs the primary interface and models reflected and transmitted radiance separately, enabling consistent radiance transport. During optimization, GLINT bootstraps transparency localization from geometry-separation cues induced by the decomposition, together with geometry and material priors from a pre-trained video relighting model. Extensive experiments demonstrate consistent improvements over prior methods for reconstructing complex transparent scenes. --- ### Consistency Beyond Contrast: Enhancing Open-Vocabulary Object Detection Robustness via Contextual Consistency Learning Authors: Bozhao Li, Shaocong Wu, Tong Shao, Senqiao Yang, Qiben Shan, Zhuotao Tian, Jingyong Su Published: 2026-03-27 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.26179v1 Recent advances in open-vocabulary object detection focus primarily on two aspects: scaling up datasets and leveraging contrastive learning to align language and vision modalities. However, these approaches often neglect internal consistency within a single modality, particularly when background or environmental changes occur. This lack of consistency leads to a performance drop because the model struggles to detect the same object in different scenes, which reveals a robustness gap. To address this issue, we introduce Contextual Consistency Learning (CCL), a novel framework that integrates two key strategies: Contextual Bootstrapped Data Generation (CBDG) and Contextual Consistency Loss (CCLoss). CBDG functions as a data generation mechanism, producing images that contain the same objects across diverse backgrounds. This is essential because existing datasets alone do not support our CCL framework. The CCLoss further enforces the invariance of object features despite environmental changes, thereby improving the model's robustness in different scenes. These strategies collectively form a unified framework for ensuring contextual consistency within the same modality. Our method achieves state-of-the-art performance, surpassing previous approaches by +16.3 AP on OmniLabel and +14.9 AP on D3. These results demonstrate the importance of enforcing intra-modal consistency, significantly enhancing model generalization in diverse environments. Our code is publicly available at: https://github.com/bozhao-li/CCL. --- ### Geometric Evolution Graph Convolutional Networks: Enhancing Graph Representation Learning via Ricci Flow Authors: Jicheng Ma, Yunyan Yang, Juan Zhao, Liang Zhao Published: 2026-03-27 Categories: cs.LG arXiv: https://arxiv.org/abs/2603.26178v2 We introduce the Geometric Evolution Graph Convolutional Network (GEGCN), a novel framework that enhances graph representation learning through explicit modeling of geometric evolution on graph structures. Specifically, GEGCN leverages a Long Short-Term Memory (LSTM) network to capture the dynamic structural sequence generated by discrete Ricci flow, and infuses the learned dynamic representations into a graph convolutional network. Extensive experiments demonstrate that GEGCN achieves excellent performance on classification tasks across various benchmark datasets, including homophilic/heterophilic graphs, filtered graphs, and large-scale graphs. --- ### Can AI Scientist Agents Learn from Lab-in-the-Loop Feedback? Evidence from Iterative Perturbation Discovery Authors: Gilles Wainrib, Barbara Bodinier, Haitem Dakhli, Josep Monserrat, Almudena Espin Perez, Sabrina Carpentier, Roberta Codato, John Klein Published: 2026-03-27 Categories: cs.LG arXiv: https://arxiv.org/abs/2603.26177v1 Recent work has questioned whether large language models (LLMs) can perform genuine in-context learning (ICL) for scientific experimental design, with prior studies suggesting that LLM-based agents exhibit no sensitivity to experimental feedback. We shed new light on this question by carrying out 800 independently replicated experiments on iterative perturbation discovery in Cell Painting high-content screening. We compare an LLM agent that iteratively updates its hypotheses using experimental feedback to a zero-shot baseline that relies solely on pretraining knowledge retrieval. Access to feedback yields a $+53.4\%$ increase in discoveries per feature on average ($p = 0.003$). To test whether this improvement arises from genuine feedback-driven learning rather than prompt-induced recall of pretraining knowledge, we introduce a random feedback control in which hit/miss labels are permuted. Under this control, the performance gain disappears, indicating that the observed improvement depends on the structure of the feedback signal ($+13.0$ hits, $p = 0.003$). We further examine how model capability affects feedback utilization. Upgrading from Claude Sonnet 4.5 to 4.6 reduces gene hallucination rates from ${\sim}33\%$--$45\%$ to ${\sim}3$--$9\%$, converting a non-significant ICL effect ($+0.8$, $p = 0.32$) into a large and highly significant improvement ($+11.0$, $p=0.003$) for the best ICL strategy. These results suggest that effective in-context learning from experimental feedback emerges only once models reach a sufficient capability threshold. --- ### CREval: An Automated Interpretable Evaluation for Creative Image Manipulation under Complex Instructions Authors: Chonghuinan Wang, Zihan Chen, Yuxiang Wei, Tianyi Jiang, Xiaohe Wu, Fan Li, Wangmeng Zuo, Hongxun Yao Published: 2026-03-27 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.26174v1 Instruction-based multimodal image manipulation has recently made rapid progress. However, existing evaluation methods lack a systematic and human-aligned framework for assessing model performance on complex and creative editing tasks. To address this gap, we propose CREval, a fully automated question-answer (QA)-based evaluation pipeline that overcomes the incompleteness and poor interpretability of opaque Multimodal Large Language Models (MLLMs) scoring. Simultaneously, we introduce CREval-Bench, a comprehensive benchmark specifically designed for creative image manipulation under complex instructions. CREval-Bench covers three categories and nine creative dimensions, comprising over 800 editing samples and 13K evaluation queries. Leveraging this pipeline and benchmark, we systematically evaluate a diverse set of state-of-the-art open and closed-source models. The results reveal that while closed-source models generally outperform open-source ones on complex and creative tasks, all models still struggle to complete such edits effectively. In addition, user studies demonstrate strong consistency between CREval's automated metrics and human judgments. Therefore, CREval provides a reliable foundation for evaluating image editing models on complex and creative image manipulation tasks, and highlights key challenges and opportunities for future research. --- ### ComVi: Context-Aware Optimized Comment Display in Video Playback Authors: Minsun Kim, Dawon Lee, Junyong Noh Published: 2026-03-27 Categories: cs.MM, cs.CV, cs.GR, cs.HC arXiv: https://arxiv.org/abs/2603.26173v1 On general video-sharing platforms like YouTube, comments are displayed independently of video playback. As viewers often read comments while watching a video, they may encounter ones referring to moments unrelated to the current scene, which can reveal spoilers and disrupt immersion. To address this problem, we present ComVi, a novel system that displays comments at contextually relevant moments, enabling viewers to see time-synchronized comments and video content together. We first map all comments to relevant video timestamps by computing audio-visual correlation, then construct the comment sequence through an optimization that considers temporal relevance, popularity (number of likes), and display duration for comfortable reading. In a user study, ComVi provided a significantly more engaging experience than conventional video interfaces (i.e., YouTube and Danmaku), with 71.9% of participants selecting ComVi as their most preferred interface. --- ### Provably Contractive and High-Quality Denoisers for Convergent Restoration Authors: Shubhi Shukla, Pravin Nair Published: 2026-03-27 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.26168v1 Image restoration, the recovery of clean images from degraded measurements, has applications in various domains like surveillance, defense, and medical imaging. Despite achieving state-of-the-art (SOTA) restoration performance, existing convolutional and attention-based networks lack stability guarantees under minor shifts in input, exposing a robustness accuracy trade-off. We develop provably contractive (global Lipschitz $< 1$) denoiser networks that considerably reduce this gap. Our design composes proximal layers obtained from unfolding techniques, with Lipschitz-controlled convolutional refinements. By contractivity, our denoiser guarantees that input perturbations of strength $\|δ\|\le\varepsilon$ induce at most $\varepsilon$ change at the output, while strong baselines such as DnCNN and Restormer can exhibit larger deviations under the same perturbations. On image denoising, the proposed model is competitive with unconstrained SOTA denoisers, reporting the tightest gap for a provably 1-Lipschitz model and establishing that such gaps are indeed achievable by contractive denoisers. Moreover, the proposed denoisers act as strong regularizers for image restoration that provably effect convergence in Plug-and-Play algorithms. Our results show that enforcing strict Lipschitz control does not inherently degrade output quality, challenging a common assumption in the literature and moving the field toward verifiable and stable vision models. Codes and pretrained models are available at https://github.com/SHUBHI1553/Contractive-Denoisers --- ### Gaussian Shannon: High-Precision Diffusion Model Watermarking Based on Communication Authors: Yi Zhang, Hongbo Huang, Liang-Jie Zhang Published: 2026-03-27 Categories: cs.CV, cs.CR arXiv: https://arxiv.org/abs/2603.26167v2 Diffusion models generate high-quality images but pose serious risks like copyright violation and disinformation. Watermarking is a key defense for tracing and authenticating AI-generated content. However, existing methods rely on threshold-based detection, which only supports fuzzy matching and cannot recover structured watermark data bit-exactly, making them unsuitable for offline verification or applications requiring lossless metadata (e.g., licensing instructions). To address this problem, in this paper, we propose Gaussian Shannon, a watermarking framework that treats the diffusion process as a noisy communication channel and enables both robust tracing and exact bit recovery. Our method embeds watermarks in the initial Gaussian noise without fine-tuning or quality loss. We identify two types of channel interference, namely local bit flips and global stochastic distortions, and design a cascaded defense combining error-correcting codes and majority voting. This ensures reliable end-to-end transmission of semantic payloads. Experiments across three Stable Diffusion variants and seven perturbation types show that Gaussian Shannon achieves state-of-the-art bit-level accuracy while maintaining a high true positive rate, enabling trustworthy rights attribution in real-world deployment. The source code have been made available at: https://github.com/Rambo-Yi/Gaussian-Shannon --- ### DataFlex: A Unified Framework for Data-Centric Dynamic Training of Large Language Models Authors: Hao Liang, Zhengyang Zhao, Meiyi Qiang, Mingrui Chen, Lu Ma, Rongyi Yu, Hengyi Feng, Shixuan Sun, Zimo Meng, Xiaochen Ma, Xuanlin Yang, Qifeng Cai et al. Published: 2026-03-27 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2603.26164v1 Data-centric training has emerged as a promising direction for improving large language models (LLMs) by optimizing not only model parameters but also the selection, composition, and weighting of training data during optimization. However, existing approaches to data selection, data mixture optimization, and data reweighting are often developed in isolated codebases with inconsistent interfaces, hindering reproducibility, fair comparison, and practical integration. In this paper, we present DataFlex, a unified data-centric dynamic training framework built upon LLaMA-Factory. DataFlex supports three major paradigms of dynamic data optimization: sample selection, domain mixture adjustment, and sample reweighting, while remaining fully compatible with the original training workflow. It provides extensible trainer abstractions and modular components, enabling a drop-in replacement for standard LLM training, and unifies key model-dependent operations such as embedding extraction, inference, and gradient computation, with support for large-scale settings including DeepSpeed ZeRO-3. We conduct comprehensive experiments across multiple data-centric methods. Dynamic data selection consistently outperforms static full-data training on MMLU across both Mistral-7B and Llama-3.2-3B. For data mixture, DoReMi and ODM improve both MMLU accuracy and corpus-level perplexity over default proportions when pretraining Qwen2.5-1.5B on SlimPajama at 6B and 30B token scales. DataFlex also achieves consistent runtime improvements over original implementations. These results demonstrate that DataFlex provides an effective, efficient, and reproducible infrastructure for data-centric dynamic training of LLMs. --- ### Clash of the models: Comparing performance of BERT-based variants for generic news frame detection Authors: Vihang Jumle Published: 2026-03-27 Categories: cs.CL, cs.CY arXiv: https://arxiv.org/abs/2603.26156v1 Framing continues to remain one of the most extensively applied theories in political communication. Developments in computation, particularly with the introduction of transformer architecture and more so with large language models (LLMs), have naturally prompted scholars to explore various novel computational approaches, especially for deductive frame detection, in recent years. While many studies have shown that different transformer models outperform their preceding models that use bag-of-words features, the debate continues to evolve regarding how these models compare with each other on classification tasks. By placing itself at this juncture, this study makes three key contributions: First, it comparatively performs generic news frame detection and compares the performance of five BERT-based variants (BERT, RoBERTa, DeBERTa, DistilBERT and ALBERT) to add to the debate on best practices around employing computational text analysis for political communication studies. Second, it introduces various fine-tuned models capable of robustly performing generic news frame detection. Third, building upon numerous previous studies that work with US-centric data, this study provides the scholarly community with a labelled generic news frames dataset based on the Swiss electoral context that aids in testing the contextual robustness of these computational approaches to framing analysis. --- ### IPV-Bench: Benchmarking Image Protection Methods under Diverse Image-to-Video Generation Scenarios Authors: Xiaofeng Li, Leyi Sheng, Yifan Zhao, Zhen Sun, Zongmin Zhang, Jiaheng Wei, Xinlei He Published: 2026-03-27 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.26154v3 Image-to-video (I2V) generation models can be misused to animate a single image into a convincing fake video, motivating perturbation-based image protection methods that aim to disrupt such generation. Yet these methods remain difficult to compare: they are reported under inconsistent metrics and generation settings, are often validated only on the single generator they were optimized against, and are evaluated on narrow, single-domain image sets that do not reflect real misuse. To address these challenges, we introduce IPV-Bench (Image Protection against Video generation), the first systematic benchmark for image protection in I2V generation scenarios. IPV-Bench couples a unified protocol that jointly scores protection effectiveness, visual fidelity, and robustness to preprocessing attacks together with IPV-500, a prompt-paired dataset spanning five misuse-relevant domains. Based on this benchmark, we evaluate five representative protection methods across four I2V models covering distinct architectures and both open-source and commercial systems. Extensive experiments show a consistently sobering picture: image protection and video disruption trade off against each other, most methods fail to disrupt generation beyond noise, protection rarely transfers across generators, and the few effective cases are broken by simple preprocessing. We further find that image content governs the perceptual cost of protection but not its benefit: no image domain offers an easier target. Overall, IPV-Bench provides a rigorous, reproducible, and extensible foundation for developing protection methods that work in practice. --- ### IPV-Bench: Benchmarking Image Protection Methods under Diverse Image-to-Video Generation Scenarios Authors: Xiaofeng Li, Leyi Sheng, Yifan Zhao, Zhen Sun, Zongmin Zhang, Jiaheng Wei, Xinlei He Published: 2026-03-27 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.26154v2 Image-to-video (I2V) generation models can be misused to animate a single image into a convincing fake video, motivating perturbation-based image protection methods that aim to disrupt such generation. Yet these methods remain difficult to compare: they are reported under inconsistent metrics and generation settings, are often validated only on the single generator they were optimized against, and are evaluated on narrow, single-domain image sets that do not reflect real misuse. To address these challenges, we introduce IPV-Bench (Image Protection against Video generation), the first systematic benchmark for image protection in I2V generation scenarios. IPV-Bench couples a unified protocol that jointly scores protection effectiveness, visual fidelity, and robustness to preprocessing attacks together with IPV-500, a prompt-paired dataset spanning five misuse-relevant domains. Based on this benchmark, we evaluate five representative protection methods across four I2V models covering distinct architectures and both open-source and commercial systems. Extensive experiments show a consistently sobering picture: image protection and video disruption trade off against each other, most methods fail to disrupt generation beyond noise, protection rarely transfers across generators, and the few effective cases are broken by simple preprocessing. We further find that image content governs the perceptual cost of protection but not its benefit: no image domain offers an easier target. Overall, IPV-Bench provides a rigorous, reproducible, and extensible foundation for developing protection methods that work in practice. --- ### Efficient Few-Shot Learning for Edge AI via Knowledge Distillation on MobileViT Authors: Shuhei Tsuyuki, Reda Bensaid, Jérémy Morlier, Mathieu Léonardon, Naoya Onizawa, Vincent Gripon, Takahiro Hanyu Published: 2026-03-27 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.26145v1 Efficient and adaptable deep learning models are an important area of deep learning research, driven by the need for highly efficient models on edge devices. Few-shot learning enables the use of deep learning models in low-data regimes, a capability that is highly sought after in real-world applications where collecting large annotated datasets is costly or impractical. This challenge is particularly relevant in edge scenarios, where connectivity may be limited, low-latency responses are required, or energy consumption constraints are critical. We propose and evaluate a pre-training method for the MobileViT backbone designed for edge computing. Specifically, we employ knowledge distillation, which transfers the generalization ability of a large-scale teacher model to a lightweight student model. This method achieves accuracy improvements of 14% and 6.7% for one-shot and five-shot classification, respectively, on the MiniImageNet benchmark, compared to the ResNet12 baseline, while reducing by 69% the number of parameters and by 88% the computational complexity of the model, in FLOPs. Furthermore, we deployed the proposed models on a Jetson Orin Nano platform and measured power consumption directly at the power supply, showing that the dynamic energy consumption is reduced by 37% with a latency of 2.6 ms. These results demonstrate that the proposed method is a promising and practical solution for deploying few-shot learning models on edge AI hardware. --- ### On the Complexity of Optimal Graph Rewiring for Oversmoothing and Oversquashing in Graph Neural Networks Authors: Mostafa Haghir Chehreghani Published: 2026-03-27 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2603.26140v1 Graph Neural Networks (GNNs) face two fundamental challenges when scaled to deep architectures: oversmoothing, where node representations converge to indistinguishable vectors, and oversquashing, where information from distant nodes fails to propagate through bottlenecks. Both phenomena are intimately tied to the underlying graph structure, raising a natural question: can we optimize the graph topology to mitigate these issues? This paper provides a theoretical investigation of the computational complexity of such graph structure optimization. We formulate oversmoothing and oversquashing mitigation as graph optimization problems based on spectral gap and conductance, respectively. We prove that exact optimization for either problem is NP-hard through reductions from Minimum Bisection, establishing NP-completeness of the decision versions. Our results provide theoretical foundations for understanding the fundamental limits of graph rewiring for GNN optimization and justify the use of approximation algorithms and heuristic methods in practice. --- ### PruneFuse: Efficient Data Selection via Weight Pruning and Network Fusion Authors: Humaira Kousar, Hasnain Irshad Bhatti, Jaekyun Moon Published: 2026-03-27 Categories: cs.LG, cs.CV arXiv: https://arxiv.org/abs/2603.26138v1 Efficient data selection is crucial for enhancing the training efficiency of deep neural networks and minimizing annotation requirements. Traditional methods often face high computational costs, limiting their scalability and practical use. We introduce PruneFuse, a novel strategy that leverages pruned networks for data selection and later fuses them with the original network to optimize training. PruneFuse operates in two stages: First, it applies structured pruning to create a smaller pruned network that, due to its structural coherence with the original network, is well-suited for the data selection task. This small network is then trained and selects the most informative samples from the dataset. Second, the trained pruned network is seamlessly fused with the original network. This integration leverages the insights gained during the training of the pruned network to facilitate the learning process of the fused network while leaving room for the network to discover more robust solutions. Extensive experimentation on various datasets demonstrates that PruneFuse significantly reduces computational costs for data selection, achieves better performance than baselines, and accelerates the overall training process. --- ### ATime-Consistent Benchmark for Repository-Level Software Engineering Evaluation Authors: Xianpeng, Sun, Haonan Sun, Tian Yu, Sheng Ma, Qincheng Zhang, Lifei Rao, Chen Tian Published: 2026-03-27 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2603.26137v1 Evaluation of repository-aware software engineering systems is often confounded by synthetic task design, prompt leakage, and temporal contamination between repository knowledge and future code changes. We present a time-consistent benchmark methodology that snapshots a repository at time T0, constructs repository-derived code knowledge using only artifacts available before T0, and evaluates on engineering tasks derived from pull requests merged in the future interval (T0, T1). Each historical pull request is transformed into a natural-language task through an LLM-assisted prompt-generation pipeline, and the benchmark is formalized as a matched A/B comparison in which the same software engineering agent is evaluated with and without repository-derived code knowledge while all other variables are held constant. We also report a baseline characterization study on two open-source repositories, DragonFly and React, using three Claude-family models and four prompt granularities. Across both repositories, file-level F1 increases monotonically from minimal to guided prompts, reaching 0.8081 on DragonFly and 0.8078 on React for the strongest tested model. These results show that prompt construction is a first-order benchmark variable. More broadly, the benchmark highlights that temporal consistency and prompt control are core validity requirements for repository-aware software engineering evaluation. --- ### PEANUT: Perturbations by Eigenvector Alignment for Attacking Graph Neural Networks Under Topology-Driven Message Passing Authors: Bhavya Kohli, Biplab Sikdar Published: 2026-03-27 Categories: cs.LG arXiv: https://arxiv.org/abs/2603.26136v3 Message Passing Neural Networks (MPNNs) have achieved strong performance on tasks involving relational data. However, small perturbations to graph structure can significantly alter their outputs, raising concerns about their robustness in real-world deployment in security critical environments. In this work, we study a core vulnerability in MPNNs that explicitly consume graph topology via the adjacency matrix or Laplacian as part of their message passing mechanism. We show that this design choice exposes an extremely vulnerable attack surface, with significant effects even under minimal perturbation. Building on this observation, we propose PEA, a simple, gradient-free, black-box injection attack that requires only a single query to the target model, capitalizing on this vulnerability by constructing a perturbation aligned with a specific significant eigenvector to induce large output deviations. Unlike graph modification attacks, PEA operates under the realistic assumption that adversaries cannot alter the original graph and are limited to injecting new nodes at inference time. PEA requires no iterative optimization, parameter learning, or surrogate models---which require additional training and remain susceptible to differences in model priors and generalization capabilities---thereby avoiding significant computational overhead and the associated transferability challenges. We evaluate PEA on popular benchmark datasets across three graph learning tasks, showing consistent performance degradation under realistic attack constraints despite its simplicity. Our results reveal a fundamental security weakness in topology-driven message passing architectures and urge an implementation shift, as the worst effects of such attacks can be substantially mitigated through appropriate input filtering. --- ### TinyML for Acoustic Anomaly Detection in IoT Sensor Networks Authors: Amar Almaini, Jakob Folz, Ghadeer Ashour Published: 2026-03-27 Categories: cs.LG arXiv: https://arxiv.org/abs/2603.26135v1 Tiny Machine Learning enables real-time, energy-efficient data processing directly on microcontrollers, making it ideal for Internet of Things sensor networks. This paper presents a compact TinyML pipeline for detecting anomalies in environmental sound within IoT sensor networks. Acoustic monitoring in IoT systems can enhance safety and context awareness, yet cloud-based processing introduces challenges related to latency, power usage, and privacy. Our pipeline addresses these issues by extracting Mel Frequency Cepstral Coefficients from sound signals and training a lightweight neural network classifier optimized for deployment on edge devices. The model was trained and evaluated using the UrbanSound8K dataset, achieving a test accuracy of 91% and balanced F1-scores of 0.91 across both normal and anomalous sound classes. These results demonstrate the feasibility and reliability of embedded acoustic anomaly detection for scalable and responsive IoT deployments. --- ### InstaVSR: Taming Diffusion for Efficient and Temporally Consistent Video Super-Resolution Authors: Jintong Hu, Bin Chen, Zhenyu Hu, Jiayue Liu, Guo Wang, Lu Qi Published: 2026-03-27 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.26134v1 Video super-resolution (VSR) seeks to reconstruct high-resolution frames from low-resolution inputs. While diffusion-based methods have substantially improved perceptual quality, extending them to video remains challenging for two reasons: strong generative priors can introduce temporal instability, and multi-frame diffusion pipelines are often too expensive for practical deployment. To address both challenges simultaneously, we propose InstaVSR, a lightweight diffusion framework for efficient video super-resolution. InstaVSR combines three ingredients: (1) a pruned one-step diffusion backbone that removes several costly components from conventional diffusion-based VSR pipelines, (2) recurrent training with flow-guided temporal regularization to improve frame-to-frame stability, and (3) dual-space adversarial learning in latent and pixel spaces to preserve perceptual quality after backbone simplification. On an NVIDIA RTX 4090, InstaVSR processes a 30-frame video at 2K$\times$2K resolution in under one minute with only 7 GB of memory usage, substantially reducing the computational cost compared to existing diffusion-based methods while maintaining favorable perceptual quality with significantly smoother temporal transitions. --- ### SWE-PRBench: Benchmarking AI Code Review Quality Against Pull Request Feedback Authors: Deepak Kumar Published: 2026-03-27 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2603.26130v1 We introduce SWE-PRBench, a benchmark of 350 pull requests with human-annotated ground truth for evaluating AI code review quality. Evaluated against an LLM-as-judge framework validated at kappa=0.75, 8 frontier models detect only 15-31% of human-flagged issues on the diff-only configuration, demonstrating that AI code review remains far below human expert performance despite strong results on code generation benchmarks. Pull requests are drawn from active open-source repositories, filtered from 700 candidates using a Repository Quality Score, and evaluated under three frozen context configurations: diff only (config_A), diff with file content (config_B), and full context (config_C), enabling systematic ablation of context provision strategies. All 8 models degrade monotonically from config_A to config_C, even when context is provided via structured semantic layers including AST-extracted function context and import graph resolution. The dominant mechanism is a collapse of Type2_Contextual issue detection at config_B, consistent with attention dilution in long contexts: a structured 2,000-token diff-with-summary prompt outperforms a 2,500-token full-context prompt enriched with execution context, behaviour mapping, and test signatures across all 8 models. The top four models are statistically indistinguishable (mean score 0.147-0.153) while a clear tier gap separates them from the remaining four (mean score <= 0.113). Dataset, contexts, annotations, and evaluation harness are released publicly. --- ### TaxaAdapter: Vision Taxonomy Models are Key to Fine-grained Image Generation over the Tree of Life Authors: Mridul Khurana, Amin Karimi Monsefi, Justin Lee, Medha Sawhney, David Carlyn, Julia Chae, Jianyang Gu, Rajiv Ramnath, Sara Beery, Wei-Lun Chao, Anuj Karpatne, Cheng Zhang Published: 2026-03-27 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.26128v1 Accurately generating images across the Tree of Life is difficult: there are over 10M distinct species on Earth, many of which differ only by subtle visual traits. Despite the remarkable progress in text-to-image synthesis, existing models often fail to capture the fine-grained visual cues that define species identity, even when their outputs appear photo-realistic. To this end, we propose TaxaAdapter, a simple and lightweight approach that incorporates Vision Taxonomy Models (VTMs) such as BioCLIP to guide fine-grained species generation. Our method injects VTM embeddings into a frozen text-to-image diffusion model, improving species-level fidelity while preserving flexible text control over attributes such as pose, style, and background. Extensive experiments demonstrate that TaxaAdapter consistently improves morphology fidelity and species-identity accuracy over strong baselines, with a cleaner architecture and training recipe. To better evaluate these improvements, we also introduce a multimodal Large Language Model-based metric that summarizes trait-level descriptions from generated and real images, providing a more interpretable measure of morphological consistency. Beyond this, we observe that TaxaAdapter exhibits strong generalization capabilities, enabling species synthesis in challenging regimes such as few-shot species with only a handful of training images and even species unseen during training. Overall, our results highlight that VTMs are a key ingredient for scalable, fine-grained species generation. --- ### Finding Distributed Object-Centric Properties in Self-Supervised Transformers Authors: Samyak Rawlekar, Amitabh Swain, Yujun Cai, Yiwei Wang, Ming-Hsuan Yang, Narendra Ahuja Published: 2026-03-27 Categories: cs.CV, cs.AI, cs.CL, cs.LG, cs.MM arXiv: https://arxiv.org/abs/2603.26127v1 Self-supervised Vision Transformers (ViTs) like DINO show an emergent ability to discover objects, typically observed in (CLS) token attention maps of the final layer. However, these maps often contain spurious activations resulting in poor localization of objects. This is because the (CLS) token, trained on an image-level objective, summarizes the entire image instead of focusing on objects. This aggregation dilutes the object-centric information existing in the local, patch-level interactions. We analyze this by computing inter-patch similarity using patch-level attention components (query, key, and value) across all layers. We find that: (1) Object-centric properties are encoded in the similarity maps derived from all three components ($q, k, v$), unlike prior work that uses only key features or the (CLS) token. (2) This object-centric information is distributed across the network, not just confined to the final layer. Based on these insights, we introduce Object-DINO, a training-free method that extracts this distributed object-centric information. Object-DINO clusters attention heads across all layers based on the similarities of their patches and automatically identifies the object-centric cluster corresponding to all objects. We demonstrate Object-DINO's effectiveness on two applications: enhancing unsupervised object discovery (+3.6 to +12.4 CorLoc gains) and mitigating object hallucination in Multimodal Large Language Models by providing visual grounding. Our results demonstrate that using this distributed object-centric information improves downstream tasks without additional training. --- ### Beyond Where to Look: Trajectory-Guided Reinforcement Learning for Multimodal RLVR Authors: Jinda Lu, Junkang Wu, Jinghan Li, Kexin Huang, Shuo Yang, Mingzhu Chen, Jiancan Wu, Kuien Liu, Xiang Wang Published: 2026-03-27 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.26126v1 Recent advances in Reinforcement Learning with Verifiable Rewards (RLVR) for multimodal large language models (MLLMs) have mainly focused on improving final answer correctness and strengthening visual grounding. However, a critical bottleneck remains: although models can attend to relevant visual regions, they often fail to effectively incorporate visual evidence into subsequent reasoning, leading to reasoning chains that are weakly grounded in visual facts. To address this issue, we propose Trajectory-Guided Reinforcement Learning (TGRL), which guides the policy model to integrate visual evidence into fine-grained reasoning processes using expert reasoning trajectories from stronger models. We further introduce token-level reweighting and trajectory filtering to ensure stable and effective policy optimization. Extensive experiments on multiple multimodal reasoning benchmarks demonstrate that TGRL consistently improves reasoning performance and effectively bridges the gap between visual perception and logical reasoning. --- ### SkinGPT-X: A Self-Evolving Collaborative Multi-Agent System for Transparent and Trustworthy Dermatological Diagnosis Authors: Zhangtianyi Chen, Yuhao Shen, Florensia Widjaja, Yan Xu, Liyuan Sun, Zijian Wang, Hongyi Chen, Wufei Dai, Juexiao Zhou Published: 2026-03-27 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2603.26122v1 While recent advancements in Large Language Models have significantly advanced dermatological diagnosis, monolithic LLMs frequently struggle with fine-grained, large-scale multi-class diagnostic tasks and rare skin disease diagnosis owing to training data sparsity, while also lacking the interpretability and traceability essential for clinical reasoning. Although multi-agent systems can offer more transparent and explainable diagnostics, existing frameworks are primarily concentrated on Visual Question Answering and conversational tasks, and their heavy reliance on static knowledge bases restricts adaptability in complex real-world clinical settings. Here, we present SkinGPT-X, a multimodal collaborative multi-agent system for dermatological diagnosis integrated with a self-evolving dermatological memory mechanism. By simulating the diagnostic workflow of dermatologists and enabling continuous memory evolution, SkinGPT-X delivers transparent and trustworthy diagnostics for the management of complex and rare dermatological cases. To validate the robustness of SkinGPT-X, we design a three-tier comparative experiment. First, we benchmark SkinGPT-X against four state-of-the-art LLMs across four public datasets, demonstrating its state-of-the-art performance with a +9.6% accuracy improvement on DDI31 and +13% weighted F1 gain on Dermnet over the state-of-the-art model. Second, we construct a large-scale multi-class dataset covering 498 distinct dermatological categories to evaluate its fine-grained classification capabilities. Finally, we curate the rare skin disease dataset, the first benchmark to address the scarcity of clinical rare skin diseases which contains 564 clinical samples with eight rare dermatological diseases. On this dataset, SkinGPT-X achieves a +9.8% accuracy improvement, a +7.1% weighted F1 improvement, a +10% Cohen's Kappa improvement. --- ### FINDER: Zero-Shot Field-Integrated Network for Distortion-free EPI Reconstruction in Diffusion MRI Authors: Namgyu Han, Seong Dae Yun, Chaeeun Lim, Sunghyun Seok, Sunju Kim, Yoonhwan Kim, Yohan Jun, Tae Hyung Kim, Berkin Bilgic, Jaejin Cho Published: 2026-03-27 Categories: eess.IV, cs.CV arXiv: https://arxiv.org/abs/2603.26117v1 Echo-planar imaging (EPI) remains the cornerstone of diffusion MRI, but it is prone to severe geometric distortions due to its rapid sampling scheme that renders the sequence highly sensitive to $B_{0}$ field inhomogeneities. While deep learning has helped improve MRI reconstruction, integrating robust geometric distortion correction into a self-supervised framework remains an unmet need. To address this, we present FINDER (Field-Integrated Network for Distortion-free EPI Reconstruction), a novel zero-shot, scan-specific framework that reformulates reconstruction as a joint optimization of the underlying image and the $B_{0}$ field map. Specifically, we employ a physics-guided unrolled network that integrates dual-domain denoisers and virtual coil extensions to enforce robust data consistency. This is coupled with an Implicit Neural Representation (INR) conditioned on spatial coordinates and latent image features to model the off-resonance field as a continuous, differentiable function. Employing an alternating minimization strategy, FINDER synergistically updates the reconstruction network and the field map, effectively disentangling susceptibility-induced geometric distortions from anatomical structures. Experimental results demonstrate that FINDER achieves superior geometric fidelity and image quality compared to state-of-the-art baselines, offering a robust solution for high-quality diffusion imaging. --- ### DPD-Cancer: Explainable Graph-Based Deep Learning for Small Molecule Anti-Cancer Activity Prediction Authors: Magnus H. Strømme, Alex G. C. de Sá, David B. Ascher Published: 2026-03-27 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2603.26114v2 DPD-Cancer is a graph-attention deep learning framework for predicting small-molecule DPD-Cancer is a graph-attention deep learning framework for predicting small-molecule anti-cancer activity across the NCI-60 panel, trained and evaluated under a strict chemistry-aware data-partitioning scheme. On the hold-out test set, the classifier achieved an Area Under the Receiver Operating Characteristic Curve (AUROC) of 0.87 (95% CI (0.86, 0.88)) and Area Under the Precision-Recall Curve (AUPRC) of 0.73 (95% CI (0.70, 0.76)); per-cell-line regression models for 73 cell lines produced a median Pearson's Correlation Coefficient (Pearson's R) of 0.64 and median Root Mean Squared Error (RMSE) of 0.67 for pGI50-value prediction. Benchmarks against pdCSM-Cancer, MLASM, and ACLPred under matched data conditions yielded consistently higher Matthew's Correlation Coefficient (MCC) scores, an occlusion-based attribution analysis confirmed that model explanations were quantitatively faithful to classifier decisions, and an applicability-domain analysis characterised reliability as a function of chemical distance. To facilitate widespread adoption, DPD-Cancer is available as a free, user-friendly web server for unrestricted use at https://biosig.lab.uq.edu.au/dpd_cancer/. --- ### SDDF: Specificity-Driven Dynamic Focusing for Open-Vocabulary Camouflaged Object Detection Authors: Jiaming Liang, Yifeng Zhan, Chunlin Liu, Weihua Zheng, Bingye Peng, Qiwei Liang, Boyang Cai, Xiaochun Mai, Qiang Nie Published: 2026-03-27 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.26109v1 Open-vocabulary object detection (OVOD) aims to detect known and unknown objects in the open world by leveraging text prompts. Benefiting from the emergence of large-scale vision--language pre-trained models, OVOD has demonstrated strong zero-shot generalization capabilities. However, when dealing with camouflaged objects, the detector often fails to distinguish and localize objects because the visual features of the objects and the background are highly similar. To bridge this gap, we construct a benchmark named OVCOD-D by augmenting carefully selected camouflaged object images with fine-grained textual descriptions. Due to the limited scale of available camouflaged object datasets, we adopt detectors pre-trained on large-scale object detection datasets as our baseline methods, as they possess stronger zero-shot generalization ability. In the specificity-aware sub-descriptions generated by multimodal large models, there still exist confusing and overly decorative modifiers. To mitigate such interference, we design a sub-description principal component contrastive fusion strategy that reduces noisy textual components. Furthermore, to address the challenge that the visual features of camouflaged objects are highly similar to those of their surrounding environment, we propose a specificity-guided regional weak alignment and dynamic focusing method, which aims to strengthen the detector's ability to discriminate camouflaged objects from background. Under the open-set evaluation setting, the proposed method achieves an AP of 56.4 on the OVCOD-D benchmark. --- ### Accurate Precipitation Forecast by Efficiently Learning from Massive Atmospheric Variables and Unbalanced Distribution Authors: Shuangliang Li, Siwei Li, Li Li, Weijie Zou, Jie Yang, Maolin Zhang Published: 2026-03-27 Categories: cs.LG, cs.CV arXiv: https://arxiv.org/abs/2603.26108v1 Short-term (0-24 hours) precipitation forecasting is highly valuable to socioeconomic activities and public safety. However, the highly complex evolution patterns of precipitation events, the extreme imbalance between precipitation and non-precipitation samples, and the inability of existing models to efficiently and effectively utilize large volumes of multi-source atmospheric observation data hinder improvements in precipitation forecasting accuracy and computational efficiency. To address the above challenges, this study developed a novel forecasting model capable of effectively and efficiently utilizing massive atmospheric observations by automatically extracting and iteratively predicting the latent features strongly associated with precipitation evolution. Furthermore, this study introduces a 'WMCE' loss function, designed to accurately discriminate extremely scarce precipitation events while precisely predicting their intensity values. Extensive experiments on two datasets demonstrate that our proposed model substantially and consistently outperforms all prevalent baselines in both accuracy and efficiency. Moreover, the proposed forecasting model substantially lowers the computational cost required to obtain valuable predictions compared to existing approaches, thereby positioning it as a milestone for efficient and practical precipitation forecasting. --- ### LLM Benchmark-User Need Misalignment for Climate Change Authors: Oucheng Liu, Lexing Xie, Jing Jiang Published: 2026-03-27 Categories: cs.CL arXiv: https://arxiv.org/abs/2603.26106v1 Climate change is a major socio-scientific issue shapes public decision-making and policy discussions. As large language models (LLMs) increasingly serve as an interface for accessing climate knowledge, whether existing benchmarks reflect user needs is critical for evaluating LLM in real-world settings. We propose a Proactive Knowledge Behaviors Framework that captures the different human-human and human-AI knowledge seeking and provision behaviors. We further develop a Topic-Intent-Form taxonomy and apply it to analyze climate-related data representing different knowledge behaviors. Our results reveal a substantial mismatch between current benchmarks and real-world user needs, while knowledge interaction patterns between humans and LLMs closely resemble those in human-human interactions. These findings provide actionable guidance for benchmark design, RAG system development, and LLM training. Code is available at https://github.com/OuchengLiu/LLM-Misalign-Climate-Change. --- ### Are LLM-Enhanced Graph Neural Networks Robust against Poisoning Attacks? Authors: Yuhang Ma, Jie Wang, Zheng Yan Published: 2026-03-27 Categories: cs.LG arXiv: https://arxiv.org/abs/2603.26105v1 Large Language Models (LLMs) have advanced Graph Neural Networks (GNNs) by enriching node representations with semantic features, giving rise to LLM-enhanced GNNs that achieve notable performance gains. However, the robustness of these models against poisoning attacks, which manipulate both graph structures and textual attributes during training, remains unexplored. To bridge this gap, we propose a robustness assessment framework that systematically evaluates LLM-enhanced GNNs under poisoning attacks. Our framework enables comprehensive evaluation across multiple dimensions. Specifically, we assess 24 victim models by combining eight LLM- or Language Model (LM)-based feature enhancers with three representative GNN backbones. To ensure diversity in attack coverage, we incorporate six structural poisoning attacks (both targeted and non-targeted) and three textual poisoning attacks operating at the character, word, and sentence levels. Furthermore, we employ four real-world datasets, including one released after the emergence of LLMs, to avoid potential ground truth leakage during LLM pretraining, thereby ensuring fair evaluation. Extensive experiments show that LLM-enhanced GNNs exhibit significantly higher accuracy and lower Relative Drop in Accuracy (RDA) than a shallow embedding-based baseline across various attack settings. Our in-depth analysis identifies key factors that contribute to this robustness, such as the effective encoding of structural and label information in node representations. Based on these insights, we outline future research directions from both offensive and defensive perspectives, and propose a new combined attack along with a graph purification defense. To support future research, we release the source code of our framework at~\url{https://github.com/CyberAlSec/LLMEGNNRP}. --- ### "Oops! ChatGPT is Temporarily Unavailable!": A Diary Study on Knowledge Workers' Experiences of LLM Withdrawal Authors: Eunseo Oh, Suyoun Lee, Jae Young Choi, Soobin Park, Youn-kyung Lim Published: 2026-03-27 Categories: cs.HC, cs.AI arXiv: https://arxiv.org/abs/2603.26099v1 LLMs have become deeply embedded in knowledge work, raising concerns about growing dependency and the potential undermining of human skills. To investigate the pervasiveness of LLMs in work practices, we conducted a four-day diary study with frequent LLM users (N=10), observing how knowledge workers responded to a temporary withdrawal of LLMs. Our findings show how LLM withdrawal disrupted participants' workflows by identifying gaps in task execution, how self-directed work led participants to reclaim professional values, and how everyday practices revealed the extent to which LLM use had become inescapably normative. Conceptualizing LLMs as infrastructural to contemporary knowledge work, this research contributes empirical insights into the often invisible role of LLMs and proposes value-driven appropriation as an approach to supporting professional values in the current LLM-pervasive work environment. --- ### A Human-Inspired Decoupled Architecture for Efficient Audio Representation Learning Authors: Harunori Kawano, Takeshi Sasaki Published: 2026-03-27 Categories: cs.SD, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2603.26098v1 While self-supervised learning (SSL) has revolutionized audio representation, the excessive parameterization and quadratic computational cost of standard Transformers limit their deployment on resource-constrained devices. To address this bottleneck, we propose HEAR (Human-inspired Efficient Audio Representation), a novel decoupled architecture. Inspired by the human cognitive ability to isolate local acoustic features from global context, HEAR splits the processing pipeline into two dedicated modules: an Acoustic Model for local feature extraction and a Task Model for global semantic integration. Coupled with an Acoustic Tokenizer trained via knowledge distillation, our approach enables robust Masked Audio Modeling (MAM). Extensive experiments demonstrate that HEAR requires only 15M parameters and 9.47 GFLOPs for inference, operating at a fraction of the computational cost of conventional foundation models (which typically require 85M-94M parameters). Despite this high efficiency, HEAR achieves highly competitive performance across diverse audio classification benchmarks. The code and pre-trained models are available at https://github.com/HarunoriKawano/HEAR --- ### Dynamic Tokenization via Reinforcement Patching: End-to-end Training and Zero-shot Transfer Authors: Yulun Wu, Sravan Kumar Ankireddy, Samuel Sharpe, Nikita Seleznev, Dehao Yuan, Hyeji Kim, Nam H. Nguyen Published: 2026-03-27 Categories: cs.LG, cs.AI, stat.ML arXiv: https://arxiv.org/abs/2603.26097v1 Efficiently aggregating spatial or temporal horizons to acquire compact representations has become a unifying principle in modern deep learning models, yet learning data-adaptive representations for long-horizon sequence data, especially continuous sequences like time series, remains an open challenge. While fixed-size patching has improved scalability and performance, discovering variable-sized, data-driven patches end-to-end often forces models to rely on soft discretization, specific backbones, or heuristic rules. In this work, we propose Reinforcement Patching (ReinPatch), the first framework to jointly optimize a sequence patching policy and its downstream sequence backbone model using reinforcement learning. By formulating patch boundary placement as a discrete decision process optimized via Group Relative Policy Gradient (GRPG), ReinPatch bypasses the need for continuous relaxations and performs dynamic patching policy optimization in a natural manner. Moreover, our method allows strict enforcement of a desired compression rate, freeing the downstream backbone to scale efficiently, and naturally supports multi-level hierarchical modeling. We evaluate ReinPatch on time-series forecasting datasets, where it demonstrates compelling performance compared to state-of-the-art data-driven patching strategies. Furthermore, our detached design allows the patching module to be extracted as a standalone foundation patcher, providing the community with visual and empirical insights into the segmentation behaviors preferred by a purely performance-driven neural patching strategy. --- ### AcTTA: Rethinking Test-Time Adaptation via Dynamic Activation Authors: Hyeongyu Kim, Geonhui Han, Dosik Hwang Published: 2026-03-27 Categories: cs.LG, cs.CV arXiv: https://arxiv.org/abs/2603.26096v1 Test-time adaptation (TTA) aims to mitigate performance degradation under distribution shifts by updating model parameters during inference. Existing approaches have primarily framed adaptation around affine modulation, focusing on recalibrating normalization layers. This perspective, while effective, overlooks another influential component in representation dynamics: the activation function. We revisit this overlooked space and propose AcTTA, an activation-aware framework that reinterprets conventional activation functions from a learnable perspective and updates them adaptively at test time. AcTTA reformulates conventional activation functions (e.g., ReLU, GELU) into parameterized forms that shift their response threshold and modulate gradient sensitivity, enabling the network to adjust activation behavior under domain shifts. This functional reparameterization enables continuous adjustment of activation behavior without modifying network weights or requiring source data. Despite its simplicity, AcTTA achieves robust and stable adaptation across diverse corruptions. Across CIFAR10-C, CIFAR100-C, and ImageNet-C, AcTTA consistently surpasses normalization-based TTA methods. Our findings highlight activation adaptation as a compact and effective route toward domain-shift-robust test-time learning, broadening the prevailing affine-centric view of adaptation. --- ### IndoBERT-Relevancy: A Context-Conditioned Relevancy Classifier for Indonesian Text Authors: Muhammad Apriandito Arya Saputra, Andry Alamsyah, Dian Puteri Ramadhani, Thomhert Suprapto Siadari, Hanif Fakhrurroja Published: 2026-03-27 Categories: cs.CL arXiv: https://arxiv.org/abs/2603.26095v1 Determining whether a piece of text is relevant to a given topic is a fundamental task in natural language processing, yet it remains largely unexplored for Bahasa Indonesia. Unlike sentiment analysis or named entity recognition, relevancy classification requires the model to reason about the relationship between two inputs simultaneously: a topical context and a candidate text. We introduce IndoBERT-Relevancy, a context-conditioned relevancy classifier built on IndoBERT Large (335M parameters) and trained on a novel dataset of 31,360 labeled pairs spanning 188 topics. Through an iterative, failure-driven data construction process, we demonstrate that no single data source is sufficient for robust relevancy classification, and that targeted synthetic data can effectively address specific model weaknesses. Our final model achieves an F1 score of 0.948 and an accuracy of 96.5%, handling both formal and informal Indonesian text. The model is publicly available at HuggingFace. --- ### CD-Buffer: Complementary Dual-Buffer Framework for Test-Time Adaptation in Adverse Weather Object Detection Authors: Youngjun Song, Hyeongyu Kim, Dosik Hwang Published: 2026-03-27 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2603.26092v1 Test-Time Adaptation (TTA) enables real-time adaptation to domain shifts without off-line retraining. Recent TTA methods have predominantly explored additive approaches that introduce lightweight modules for feature refinement. Recently, a subtractive approach that removes domain-sensitive channels has emerged as an alternative direction. We observe that these paradigms exhibit complementary effectiveness patterns: subtractive methods excel under severe shifts by removing corrupted features, while additive methods are effective under moderate shifts requiring refinement. However, each paradigm operates effectively only within limited shift severity ranges, failing to generalize across diverse corruption levels. This leads to the following question: can we adaptively balance both strategies based on measured feature-level domain shift? We propose CD-Buffer, a novel complementary dual-buffer framework where subtractive and additive mechanisms operate in opposite yet coordinated directions driven by a unified discrepancy metric. Our key innovation lies in the discrepancy-driven coupling: Our framework couples removal and refinement through a unified discrepancy metric, automatically balancing both strategies based on feature-level shift severity. This establishes automatic channel-wise balancing that adapts differentiated treatment to heterogeneous shift magnitudes without manual tuning. Extensive experiments on KITTI, Cityscapes, and ACDC datasets demonstrate state-of-the-art performance, consistently achieving superior results across diverse weather conditions and severity levels. --- ### Selective Deficits in LLM Mental Self-Modeling in a Behavior-Based Test of Theory of Mind Authors: Christopher Ackerman Published: 2026-03-27 Categories: cs.LG, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2603.26089v3 The ability to represent oneself and others as agents with knowledge, intentions, and belief states that guide their behavior - Theory of Mind - is a human universal that enables us to navigate - and manipulate - the social world. It is supported by our ability to form mental models of ourselves and others. Its ubiquity in human affairs entails that LLMs have seen innumerable examples of it in their training data and therefore may have learned to mimic it, but whether they have actually learned causal models that they can deploy in arbitrary settings is unclear. We therefore develop a novel experimental paradigm that requires that subjects form representations of the mental states of themselves and others and act on them strategically rather than merely describe them. We test a wide range of leading open and closed source LLMs released since 2024, as well as human subjects, on this paradigm. We find that 1) LLMs released before mid-2025 fail at all of our tasks, 2) more recent LLMs achieve human-level performance on modeling the cognitive states of others, and 3) even frontier LLMs fail at our self-modeling task - unless afforded a scratchpad in the form of a reasoning trace. We further demonstrate cognitive load effects on other-modeling tasks, offering suggestive evidence that LLMs are using something akin to limited-capacity working memory to hold these mental representations in mind during a single forward pass. Finally, we explore the mechanisms by which reasoning models succeed at the self- and other-modeling tasks, and show that they readily engage in strategic deception. --- ### Selective Deficits in LLM Mental Self-Modeling in a Behavior-Based Test of Theory of Mind Authors: Christopher Ackerman Published: 2026-03-27 Categories: cs.LG, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2603.26089v2 The ability to represent oneself and others as agents with knowledge, intentions, and belief states that guide their behavior - Theory of Mind - is a human universal that enables us to navigate - and manipulate - the social world. It is supported by our ability to form mental models of ourselves and others. Its ubiquity in human affairs entails that LLMs have seen innumerable examples of it in their training data and therefore may have learned to mimic it, but whether they have actually learned causal models that they can deploy in arbitrary settings is unclear. We therefore develop a novel experimental paradigm that requires that subjects form representations of the mental states of themselves and others and act on them strategically rather than merely describe them. We test a wide range of leading open and closed source LLMs released since 2024, as well as human subjects, on this paradigm. We find that 1) LLMs released before mid-2025 fail at all of our tasks, 2) more recent LLMs achieve human-level performance on modeling the cognitive states of others, and 3) even frontier LLMs fail at our self-modeling task - unless afforded a scratchpad in the form of a reasoning trace. We further demonstrate cognitive load effects on other-modeling tasks, offering suggestive evidence that LLMs are using something akin to limited-capacity working memory to hold these mental representations in mind during a single forward pass. Finally, we explore the mechanisms by which reasoning models succeed at the self- and other-modeling tasks, and show that they readily engage in strategic deception. --- ### Learnable Instance Attention Filtering for Adaptive Detector Distillation Authors: Chen Liu, Qizhen Lan, Zhicheng Ding, Xinyu Chu, Qing Tian Published: 2026-03-27 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.26088v1 As deep vision models grow increasingly complex to achieve higher performance, deployment efficiency has become a critical concern. Knowledge distillation (KD) mitigates this issue by transferring knowledge from large teacher models to compact student models. While many feature-based KD methods rely on spatial filtering to guide distillation, they typically treat all object instances uniformly, ignoring instance-level variability. Moreover, existing attention filtering mechanisms are typically heuristic or teacher-driven, rather than learned with the student. To address these limitations, we propose Learnable Instance Attention Filtering for Adaptive Detector Distillation (LIAF-KD), a novel framework that introduces learnable instance selectors to dynamically evaluate and reweight instance importance during distillation. Notably, the student contributes to this process based on its evolving learning state. Experiments on the KITTI and COCO datasets demonstrate consistent improvements, with a 2% gain on a GFL ResNet-50 student without added complexity, outperforming state-of-the-art methods. --- ### Experimental study on surveillance video-based indoor occupancy measurement with occupant-centric control Authors: Irfan Qaisar, Kailai Sun, Qingshan Jia, Qianchuan Zhao Published: 2026-03-27 Categories: eess.SY, cs.CV arXiv: https://arxiv.org/abs/2603.26081v1 Accurate occupancy information is essential for closed-loop occupant-centric control (OCC) in smart buildings. However, existing vision-based occupancy measurement methods often struggle to provide stable and accurate measurements in real indoor environments, and their implications for downstream HVAC control remain insufficiently studied. To achieve Net Zero emissions by 2050, this paper presents an experimental study of large language models (LLMs)-enhanced vision-based indoor occupancy measurement and its impact on OCC-enabled HVAC operation. Detection-only, tracking-based, and LLM-based refinement pipelines are compared under identical conditions using real surveillance data collected from a research laboratory in China, with frame-level manual ground-truth annotations. Results show that tracking-based methods improve temporal stability over detection-only measurement, while LLM-based refinement further improves occupancy measurement performance and reduces false unoccupied prediction. The best-performing pipeline, YOLOv8+DeepSeek, achieves an accuracy of 0.8824 and an F1-score of 0.9320. This pipeline is then integrated into an HVAC supervisory model predictive control framework in OpenStudio-EnergyPlus. Experimental results demonstrate that the proposed framework can support more efficient OCC operation, achieving a substantial HVAC energy-saving potential of 17.94%. These findings provide an effective methodology and practical foundation for future research in AI-enhanced smart building operations. --- ### When Identities Collapse: A Stress-Test Benchmark for Multi-Subject Personalization Authors: Zhihan Chen, Yuhuan Zhao, Yijie Zhu, Xinyu Yao Published: 2026-03-27 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2603.26078v1 Subject-driven text-to-image diffusion models have achieved remarkable success in preserving single identities, yet their ability to compose multiple interacting subjects remains largely unexplored and highly challenging. Existing evaluation protocols typically rely on global CLIP metrics, which are insensitive to local identity collapse and fail to capture the severity of multi-subject entanglement. In this paper, we identify a pervasive "Illusion of Scalability" in current models: while they excel at synthesizing 2-4 subjects in simple layouts, they suffer from catastrophic identity collapse when scaled to 6-10 subjects or tasked with complex physical interactions. To systematically expose this failure mode, we construct a rigorous stress-test benchmark comprising 75 prompts distributed across varying subject counts and interaction difficulties (Neutral, Occlusion, Interaction). Furthermore, we demonstrate that standard CLIP-based metrics are fundamentally flawed for this task, as they often assign high scores to semantically correct but identity-collapsed images (e.g., generating generic clones). To address this, we introduce the Subject Collapse Rate (SCR), a novel evaluation metric grounded in DINOv2's structural priors, which strictly penalizes local attention leakage and homogenization. Our extensive evaluation of state-of-the-art models (MOSAIC, XVerse, PSR) reveals a precipitous drop in identity fidelity as scene complexity grows, with SCR approaching 100% at 10 subjects. We trace this collapse to the semantic shortcuts inherent in global attention routing, underscoring the urgent need for explicit physical disentanglement in future generative architectures. --- ### Semi-Automated Knowledge Engineering and Process Mapping for Total Airport Management Authors: Darryl Teo, Adharsha Sam, Chuan Shen Marcus Koh, Rakesh Nagi, Nuno Antunes Ribeiro Published: 2026-03-27 Categories: cs.AI, cs.CL, cs.IR arXiv: https://arxiv.org/abs/2603.26076v1 Documentation of airport operations is inherently complex due to extensive technical terminology, rigorous regulations, proprietary regional information, and fragmented communication across multiple stakeholders. The resulting data silos and semantic inconsistencies present a significant impediment to the Total Airport Management (TAM) initiative. This paper presents a methodological framework for constructing a domain-grounded, machine-readable Knowledge Graph (KG) through a dual-stage fusion of symbolic Knowledge Engineering (KE) and generative Large Language Models (LLMs). The framework employs a scaffolded fusion strategy in which expert-curated KE structures guide LLM prompts to facilitate the discovery of semantically aligned knowledge triples. We evaluate this methodology on the Google LangExtract library and investigate the impact of context window utilization by comparing localized segment-based inference with document-level processing. Contrary to prior empirical observations of long-context degradation in LLMs, document-level processing improves the recovery of non-linear procedural dependencies. To ensure the high-fidelity provenance required in airport operations, the proposed framework fuses a probabilistic model for discovery and a deterministic algorithm for anchoring every extraction to its ground source. This ensures absolute traceability and verifiability, bridging the gap between "black-box" generative outputs and the transparency required for operational tooling. Finally, we introduce an automated framework that operationalizes this pipeline to synthesize complex operational workflows from unstructured textual corpora. --- ### MUST: Modality-Specific Representation-Aware Transformer for Diffusion-Enhanced Survival Prediction with Missing Modality Authors: Kyungwon Kim, Dosik Hwang Published: 2026-03-27 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2603.26071v1 Accurate survival prediction from multimodal medical data is essential for precision oncology, yet clinical deployment faces a persistent challenge: modalities are frequently incomplete due to cost constraints, technical limitations, or retrospective data availability. While recent methods attempt to address missing modalities through feature alignment or joint distribution learning, they fundamentally lack explicit modeling of the unique contributions of each modality as opposed to the information derivable from other modalities. We propose MUST (Modality-Specific representation-aware Transformer), a novel framework that explicitly decomposes each modality's representation into modality-specific and cross-modal contextualized components through algebraic constraints in a learned low-rank shared subspace. This decomposition enables precise identification of what information is lost when a modality is absent. For the truly modality-specific information that cannot be inferred from available modalities, we employ conditional latent diffusion models to generate high-quality representations conditioned on recovered shared information and learned structural priors. Extensive experiments on five TCGA cancer datasets demonstrate that MUST achieves state-of-the-art performance with complete data while maintaining robust predictions in both missing pathology and missing genomics conditions, with clinically acceptable inference latency. --- ### PAD-Hand: Physics-Aware Diffusion for Hand Motion Recovery Authors: Elkhan Ismayilzada, Yufei Zhang, Zijun Cui Published: 2026-03-27 Categories: cs.CV arXiv: https://arxiv.org/abs/2603.26068v2 Significant advancements made in reconstructing hands from images have delivered accurate single-frame estimates, yet they often lack physics consistency and provide no notion of how confidently the motion satisfies physics. In this paper, we propose a novel physics-aware conditional diffusion framework that refines noisy pose sequences into physically plausible hand motion while estimating the physics variance in motion estimates. Building on a MeshCNN-Transformer backbone, we formulate Euler-Lagrange dynamics for articulated hands. Unlike prior works that enforce zero residuals, we treat the resulting dynamic residuals as virtual observables to more effectively integrate physics. Through a last-layer Laplace approximation, our method produces per-joint, per-time variances that measure physics consistency and offers interpretable variance maps indicating where physical consistency weakens. Experiments on two well-known hand datasets show consistent gains over strong image-based initializations and competitive video-based methods. Qualitative results confirm that our variance estimations are aligned with the physical plausibility of the motion in image-based estimates. --- ### R-PGA: Robust Physical Adversarial Camouflage Generation via Relightable 3D Gaussian Splatting Authors: Tianrui Lou, Siyuan Liang, Jiawei Liang, Yuze Gao, Xiaochun Cao Published: 2026-03-27 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2603.26067v1 Physical adversarial camouflage poses a severe security threat to autonomous driving systems by mapping adversarial textures onto 3D objects. Nevertheless, current methods remain brittle in complex dynamic scenarios, failing to generalize across diverse geometric (e.g., viewing configurations) and radiometric (e.g., dynamic illumination, atmospheric scattering) variations. We attribute this deficiency to two fundamental limitations in simulation and optimization. First, the reliance on coarse, oversimplified simulations (e.g., via CARLA) induces a significant domain gap, confining optimization to a biased feature space. Second, standard strategies targeting average performance result in a rugged loss landscape, leaving the camouflage vulnerable to configuration shifts.To bridge these gaps, we propose the Relightable Physical 3D Gaussian Splatting (3DGS) based Attack framework (R-PGA). Technically, to address the simulation fidelity issue, we leverage 3DGS to ensure photo-realistic reconstruction and augment it with physically disentangled attributes to decouple intrinsic material from lighting. Furthermore, we design a hybrid rendering pipeline that leverages precise Relightable 3DGS for foreground rendering, while employing a pre-trained image translation model to synthesize plausible relighted backgrounds that align with the relighted foreground.To address the optimization robustness issue, we propose the Hard Physical Configuration Mining (HPCM) module, designed to actively mine worst-case physical configurations and suppress their corresponding loss peaks. This strategy not only diminishes the overall loss magnitude but also effectively flattens the rugged loss landscape, ensuring consistent adversarial effectiveness and robustness across varying physical configurations. ---