How do LLMs Compute Verbal Confidence

Verbal confidence -- prompting LLMs to state their confidence as a number or category -- is widely used to extract uncertainty estimates from black-box models. However, how LLMs internally generate such scores remains unknown. We address two questions: first, when confidence is computed - just-in-time when requested, or automatically during answer generation and cached for later retrieval; and second, what verbal confidence represents - token log-probabilities, or a richer evaluation of answer quality? Focusing on Gemma 3 27B and Qwen 2.5 7B, we provide convergent evidence for cached retrieval. Activation steering, patching, noising, and swap experiments reveal that confidence representations emerge at answer-adjacent positions before appearing at the verbalization site. Attention blocking pinpoints the information flow: confidence is gathered from answer tokens, cached at the first post-answer position, then retrieved for output. Critically, linear probing and variance partitioning reveal that these cached representations explain substantial variance in verbal confidence beyond token log-probabilities, suggesting a richer answer-quality evaluation rather than a simple fluency readout. These findings demonstrate that verbal confidence reflects automatic, sophisticated self-evaluation -- not post-hoc reconstruction -- with implications for understanding metacognition in LLMs and improving calibration.

WebPII: Benchmarking Visual PII Detection for Computer-Use Agents

Computer use agents create new privacy risks: training data collected from real websites inevitably contains sensitive information, and cloud-hosted inference exposes user screenshots. Detecting personally identifiable information in web screenshots is critical for privacy-preserving deployment, but no public benchmark exists for this task. We introduce WebPII, a fine-grained synthetic benchmark of 44,865 annotated e-commerce UI images designed with three key properties: extended PII taxonomy including transaction-level identifiers that enable reidentification, anticipatory detection for partially-filled forms where users are actively entering data, and scalable generation through VLM-based UI reproduction. Experiments validate that these design choices improve layout-invariant detection across diverse interfaces and generalization to held-out page types. We train WebRedact to demonstrate practical utility, more than doubling text-extraction baseline accuracy (0.753 vs 0.357 mAP@50) at real-time CPU latency (20ms). We release the dataset and model to support privacy-preserving computer use research.

EPOFusion: Exposure aware Progressive Optimization Method for Infrared and Visible Image Fusion

Overexposure frequently occurs in practical scenarios, causing the loss of critical visual information. However, existing infrared and visible fusion methods still exhibit unsatisfactory performance in highly bright regions. To address this, we propose EPOFusion, an exposure-aware fusion model. Specifically, a guidance module is introduced to facilitate the encoder in extracting fine-grained infrared features from overexposed regions. Meanwhile, an iterative decoder incorporating a multiscale context fusion module is designed to progressively enhance the fused image, ensuring consistent details and superior visual quality. Finally, an adaptive loss function dynamically constrains the fusion process, enabling an effective balance between the modalities under varying exposure conditions. To achieve better exposure awareness, we construct the first infrared and visible overexposure dataset (IVOE) with high quality infrared guided annotations for overexposed regions. Extensive experiments show that EPOFusion outperforms existing methods. It maintains infrared cues in overexposed regions while achieving visually faithful fusion in non-overexposed areas, thereby enhancing both visual fidelity and downstream task performance. Code, fusion results and IVOE dataset will be made available at https://github.com/warren-wzw/EPOFusion.git.

Multimodal Connectome Fusion via Cross-Attention for Autism Spectrum Disorder Classification Using Graph Learning

Autism spectrum disorder (ASD) is a complex neurodevelopmental condition characterized by atypical functional brain connectivity and subtle structural alterations. rs-fMRI has been widely used to identify disruptions in large-scale brain networks, while structural MRI provides complementary information about morphological organization. Despite their complementary nature, effectively integrating these heterogeneous imaging modalities within a unified framework remains challenging. This study proposes a multimodal graph learning framework that preserves the dominant role of functional connectivity while integrating structural imaging and phenotypic information for ASD classification. The proposed framework is evaluated on ABIDE-I dataset. Each subject is represented as a node within a population graph. Functional and structural features are extracted as modality-specific node attributes, while inter-subject relationships are modeled using a pairwise association encoder (PAE) based on phenotypic information. Two Edge Variational GCNs are trained to learn subject-level embeddings. To enable effective multimodal integration, we introduce a novel asymmetric transformer-based cross-attention mechanism that allows functional embeddings to selectively incorporate complementary structural information while preserving functional dominance. The fused embeddings are then passed to a MLP for ASD classification. Using stratified 10-fold cross-validation, the framework achieved an AUC of 87.3% and an accuracy of 84.4%. Under leave-one-site-out cross-validation (LOSO-CV), the model achieved an average cross-site accuracy of 82.0%, outperforming existing methods by approximately 3% under 10-fold cross-validation and 7% under LOSO-CV. The proposed framework effectively integrates heterogeneous multimodal data from the multi-site ABIDE-I dataset, improving automated ASD classification across imaging sites.

Preliminary analysis of RGB-NIR Image Registration techniques for off-road forestry environments

RGB-NIR image registration plays an important role in sensor-fusion, image enhancement and off-road autonomy. In this work, we evaluate both classical and Deep Learning (DL) based image registration techniques to access their suitability for off-road forestry applications. NeMAR, trained under 6 different configurations, demonstrates partial success however, its GAN loss instability suggests challenges in preserving geometric consistency. MURF, when tested on off-road forestry data shows promising large scale feature alignment during shared information extraction but struggles with fine details in dense vegetation. Even though this is just a preliminary evaluation, our study necessitates further refinements for robust, multi-scale registration for off-road forest applications.

Learning from Limited and Incomplete Data: A Multimodal Framework for Predicting Pathological Response in NSCLC

Major pathological response (pR) following neoadjuvant therapy is a clinically meaningful endpoint in non-small cell lung cancer, strongly associated with improved survival. However, accurate preoperative prediction of pR remains challenging, particularly in real-world clinical settings characterized by limited data availability and incomplete clinical profiles. In this study, we propose a multimodal deep learning framework designed to address these constraints by integrating foundation model-based CT feature extraction with a missing-aware architecture for clinical variables. This approach enables robust learning from small cohorts while explicitly modeling missing clinical information, without relying on conventional imputation strategies. A weighted fusion mechanism is employed to leverage the complementary contributions of imaging and clinical modalities, yielding a multimodal model that consistently outperforms both unimodal imaging and clinical baselines. These findings underscore the added value of integrating heterogeneous data sources and highlight the potential of multimodal, missing-aware systems to support pR prediction under realistic clinical conditions.

Performance Evaluation of Open-Source Large Language Models for Assisting Pathology Report Writing in Japanese

The performance of large language models (LLMs) for supporting pathology report writing in Japanese remains unexplored. We evaluated seven open-source LLMs from three perspectives: (A) generation and information extraction of pathology diagnosis text following predefined formats, (B) correction of typographical errors in Japanese pathology reports, and (C) subjective evaluation of model-generated explanatory text by pathologists and clinicians. Thinking models and medical-specialized models showed advantages in structured reporting tasks that required reasoning and in typo correction. In contrast, preferences for explanatory outputs varied substantially across raters. Although the utility of LLMs differed by task, our findings suggest that open-source LLMs can be useful for assisting Japanese pathology report writing in limited but clinically relevant scenarios.

Halfway to 3D: Ensembling 2.5D and 3D Models for Robust COVID-19 CT Diagnosis

We propose a deep learning framework for COVID-19 detection and disease classification from chest CT scans that integrates both 2.5D and 3D representations to capture complementary slice-level and volumetric information. The 2.5D branch processes multi-view CT slices (axial, coronal, sagittal) using a DINOv3 vision transformer to extract robust visual features, while the 3D branch employs a ResNet-18 architecture to model volumetric context and is pretrained with Variance Risk Extrapolation (VREx) followed by supervised contrastive learning to improve cross-source robustness. Predictions from both branches are combined through logit-level ensemble inference. Experiments on the PHAROS-AIF-MIH benchmark demonstrate the effectiveness of the proposed approach: for binary COVID-19 detection, the ensemble achieves 94.48% accuracy and a 0.9426 Macro F1-score, outperforming both individual models, while for multi-class disease classification the 2.5D DINOv3 model achieves the best performance with 79.35% accuracy and a 0.7497 Macro F1-score. These results highlight the benefit of combining pretrained slice-based representations with volumetric modeling for robust multi-source medical imaging analysis. Code is available at https://github.com/HySonLab/PHAROS-AIF-MIH

GoalSwarm: Multi-UAV Semantic Coordination for Open-Vocabulary Object Navigation

Cooperative visual semantic navigation is a foundational capability for aerial robot teams operating in unknown environments. However, achieving robust open-vocabulary object-goal navigation remains challenging due to the computational constraints of deploying heavy perception models onboard and the complexity of decentralized multi-agent coordination. We present GoalSwarm, a fully decentralized multi-UAV framework for zero-shot semantic object-goal navigation. Each UAV collaboratively constructs a shared, lightweight 2D top-down semantic occupancy map by projecting depth observations from aerial vantage points, eliminating the computational burden of full 3D representations while preserving essential geometric and semantic structure. The core contributions of GoalSwarm are threefold: (1) integration of zero-shot foundation model -- SAM3 for open vocabulary detection and pixel-level segmentation, enabling open-vocabulary target identification without task-specific training; (2) a Bayesian Value Map that fuses multi-viewpoint detection confidences into a per-pixel goal-relevance distribution, enabling informed frontier scoring via Upper Confidence Bound (UCB) exploration; and (3) a decentralized coordination strategy combining semantic frontier extraction, cost-utility bidding with geodesic path costs, and spatial separation penalties to minimize redundant exploration across the swarm.

InViC: Intent-aware Visual Cues for Medical Visual Question Answering

Medical visual question answering (Med-VQA) aims to answer clinically relevant questions grounded in medical images. However, existing multimodal large language models (MLLMs) often exhibit shortcut answering, producing plausible responses by exploiting language priors or dataset biases while insufficiently attending to visual evidence. This behavior undermines clinical reliability, especially when subtle imaging findings are decisive. We propose a lightweight plug-in framework, termed Intent-aware Visual Cues (InViC), to explicitly enhance image-based answer generation in medical VQA. InViC introduces a Cue Tokens Extraction (CTE) module that distills dense visual tokens into a compact set of K question-conditioned cue tokens, which serve as structured visual intermediaries injected into the LLM decoder to promote intent-aligned visual evidence. To discourage bypassing of visual information, we further design a two-stage fine-tuning strategy with a cue-bottleneck attention mask. In Stage I, we employ an attention mask to block the LLM's direct view of raw visual features, thereby funneling all visual evidence through the cue pathway. In Stage II, standard causal attention is restored to train the LLM to jointly exploit the visual and cue tokens. We evaluate InViC on three public Med-VQA benchmarks (VQA-RAD, SLAKE, and ImageCLEF VQA-Med 2019) across multiple representative MLLMs. InViC consistently improves over zero-shot inference and standard LoRA fine-tuning, demonstrating that intent-aware visual cues with bottlenecked training is a practical and effective strategy for improving trustworthy Med-VQA.