# Research Papers — 2025-11 · bundle page 108 of 197 > Full-text markdown of papers #5351–5400 of 9830 published in 2025-11, newest first. AI-readable mirrors of the full 9830-paper 2025-11 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2025-11/page/108 [← Previous bundle](https://aipod.dev/papers/archive/2025-11/page/107.md) [Next bundle →](https://aipod.dev/papers/archive/2025-11/page/109.md) ### Robustness of LLM-enabled vehicle trajectory prediction under data security threats Authors: Feilong Wang, Fuqiang Liu Published: 2025-11-14 Categories: cs.LG, cs.AI, cs.CR arXiv: https://arxiv.org/abs/2511.13753v1 The integration of large language models (LLMs) into automated driving systems has opened new possibilities for reasoning and decision-making by transforming complex driving contexts into language-understandable representations. Recent studies demonstrate that fine-tuned LLMs can accurately predict vehicle trajectories and lane-change intentions by gathering and transforming data from surrounding vehicles. However, the robustness of such LLM-based prediction models for safety-critical driving systems remains unexplored, despite the increasing concerns about the trustworthiness of LLMs. This study addresses this gap by conducting a systematic vulnerability analysis of LLM-enabled vehicle trajectory prediction. We propose a one-feature differential evolution attack that perturbs a single kinematic feature of surrounding vehicles within the LLM's input prompts under a black-box setting. Experiments on the highD dataset reveal that even minor, physically plausible perturbations can significantly disrupt model outputs, underscoring the susceptibility of LLM-based predictors to adversarial manipulation. Further analyses reveal a trade-off between accuracy and robustness, examine the failure mechanism, and explore potential mitigation solutions. The findings provide the very first insights into adversarial vulnerabilities of LLM-driven automated vehicle models in the context of vehicular interactions and highlight the need for robustness-oriented design in future LLM-based intelligent transportation systems. --- ### Motor Imagery Classification Using Feature Fusion of Spatially Weighted Electroencephalography Authors: Abdullah Al Shiam, Md. Khademul Islam Molla, Abu Saleh Musa Miah, Md. Abdus Samad Kamal Published: 2025-11-14 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2511.13752v1 A Brain Computer Interface (BCI) connects the human brain to the outside world, providing a direct communication channel. Electroencephalography (EEG) signals are commonly used in BCIs to reflect cognitive patterns related to motor function activities. However, due to the multichannel nature of EEG signals, explicit information processing is crucial to lessen computational complexity in BCI systems. This study proposes an innovative method based on brain region-specific channel selection and multi-domain feature fusion to improve classification accuracy. The novelty of the proposed approach lies in region-based channel selection, where EEG channels are grouped according to their functional relevance to distinct brain regions. By selecting channels based on specific regions involved in motor imagery (MI) tasks, this technique eliminates irrelevant channels, reducing data dimensionality and improving computational efficiency. This also ensures that the extracted features are more reflective of the brain actual activity related to motor tasks. Three distinct feature extraction methods Common Spatial Pattern (CSP), Fuzzy C-means clustering, and Tangent Space Mapping (TSM), are applied to each group of channels based on their brain region. Each method targets different characteristics of the EEG signal: CSP focuses on spatial patterns, Fuzzy C means identifies clusters within the data, and TSM captures non-linear patterns in the signal. The combined feature vector is used to classify motor imagery tasks (left hand, right hand, and right foot) using Support Vector Machine (SVM). The proposed method was validated on publicly available benchmark EEG datasets (IVA and I) from the BCI competition III and IV. The results show that the approach outperforms existing methods, achieving classification accuracies of 90.77% and 84.50% for datasets IVA and I, respectively. --- ### Temporal Micro-Doppler Spectrogram-based ViT Multiclass Target Classification Authors: Nghia Thinh Nguyen, Tri Nhu Do Published: 2025-11-14 Categories: eess.SP, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2511.11951v1 In this paper, we propose a new Temporal MDS-Vision Transformer (T-MDS-ViT) for multiclass target classification using millimeter-wave FMCW radar micro-Doppler spectrograms. Specifically, we design a transformer-based architecture that processes stacked range-velocity-angle (RVA) spatiotemporal tensors via patch embeddings and cross-axis attention mechanisms to explicitly model the sequential nature of MDS data across multiple frames. The T-MDS-ViT exploits mobility-aware constraints in its attention layer correspondences to maintain separability under target overlaps and partial occlusions. Next, we apply an explainable mechanism to examine how the attention layers focus on characteristic high-energy regions of the MDS representations and their effect on class-specific kinematic features. We also demonstrate that our proposed framework is superior to existing CNN-based methods in terms of classification accuracy while achieving better data efficiency and real-time deployability. --- ### Computation-aware Energy-harvesting Federated Learning with Pipelined Cyclic Scheduling Authors: Eunjeong Jeong, Nikolaos Pappas Published: 2025-11-14 Categories: cs.LG, cs.IT arXiv: https://arxiv.org/abs/2511.11949v2 Federated learning (FL) is a powerful paradigm for distributed learning, but increasing model complexity leads to significant energy consumption from client-side computations for local training. This challenge is critical in energy-harvesting FL (EHFL) systems, where the participation availability of each device fluctuates because of limited energy. To address this, we propose PipeCycle, a battery-aware distributed learning framework that organizes clients into pipelined cyclic groups. When a group completes its intra-group aggregation, its aggregated model is relayed directly to a newly formed group as a reference for local training, allowing multiple groups to coexist in the pipeline while overlapping client recharging periods with active training in other pipeline stages. We provide a convergence analysis of PipeCycle under a realistic energy consumption model in which local training spans multiple time slots, and show that the cyclic structure of the pipeline imposes a finite-horizon staleness bound that avoids the exponential factors typical of asynchronous FL analyses. Numerical experiments across both IID and non-IID data and various battery charging probabilities show that PipeCycle reaches a target accuracy with substantially lower cumulative energy than existing FL baselines, particularly under severe label skew where competing cyclic schemes collapse to near-chance accuracy. --- ### AI-Open-RAN for Non-Terrestrial Networks Authors: Tri Nhu Do Published: 2025-11-14 Categories: eess.SP, cs.AI arXiv: https://arxiv.org/abs/2511.11947v1 In this paper, we propose the concept of AIO-RAN-NTN, a unified all-in-one Radio Access Network (RAN) for Non-Terrestrial Networks (NTNs), built on an open architecture that leverages open interfaces and artificial intelligence (AI)-based functionalities. This approach advances interoperability, flexibility, and intelligence in next-generation telecommunications. First, we provide a concise overview of the state-of-the-art architectures for Open-RAN and AI-RAN, highlighting key network functions and infrastructure elements. Next, we introduce our integrated AIO-RAN-NTN blueprint, emphasizing how internal and air interfaces from AIO-RAN and the 3rd Generation Partnership Project (3GPP) can be applied to emerging environments such as NTNs. To examine the impact of mobility on AIO-RAN, we implement a testbed transmission using the OpenAirInterface platform for a standalone (SA) New Radio (NR) 5G system. We then train an AI model on realistic data to forecast key performance indicators (KPIs). Our experiments demonstrate that the AIO-based SA architecture is sensitive to mobility, even at low speeds, but this limitation can be mitigated through AI-driven KPI forecasting. --- ### Improving LLM's Attachment to External Knowledge In Dialogue Generation Tasks Through Entity Anonymization Authors: Hadi Sheikhi, Chenyang Huang, Osmar R. Zaïane Published: 2025-11-14 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2511.11946v1 Knowledge graph-based dialogue generation (KG-DG) is a challenging task requiring models to effectively incorporate external knowledge into conversational responses. While large language models (LLMs) have achieved impressive results across various NLP tasks, their ability to utilize external knowledge in KG-DG remains under-explored. We observe that LLMs often rely on internal knowledge, leading to detachment from provided knowledge graphs, even when they are given a flawlessly retrieved knowledge graph. First, we introduce LLM-KAT, an evaluation procedure for measuring knowledge attachment in generated responses. Second, we propose a simple yet effective entity anonymization technique to encourage LLMs to better leverage external knowledge. Experiments on the OpenDialKG dataset demonstrate that our approach improves LLMs' attachment on external knowledge. --- ### Augmenting The Weather: A Hybrid Counterfactual-SMOTE Algorithm for Improving Crop Growth Prediction When Climate Changes Authors: Mohammed Temraz, Mark T Keane Published: 2025-11-14 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2511.11945v1 In recent years, humanity has begun to experience the catastrophic effects of climate change as economic sectors (such as agriculture) struggle with unpredictable and extreme weather events. Artificial Intelligence (AI) should help us handle these climate challenges but its most promising solutions are not good at dealing with climate-disrupted data; specifically, machine learning methods that work from historical data-distributions, are not good at handling out-of-distribution, outlier events. In this paper, we propose a novel data augmentation method, that treats the predictive problems around climate change as being, in part, due to class-imbalance issues; that is, prediction from historical datasets is difficult because, by definition, they lack sufficient minority-class instances of "climate outlier events". This novel data augmentation method -- called Counterfactual-Based SMOTE (CFA-SMOTE) -- combines an instance-based counterfactual method from Explainable AI (XAI) with the well-known class-imbalance method, SMOTE. CFA-SMOTE creates synthetic data-points representing outlier, climate-events that augment the dataset to improve predictive performance. We report comparative experiments using this CFA-SMOTE method, comparing it to benchmark counterfactual and class-imbalance methods under different conditions (i.e., class-imbalance ratios). The focal climate-change domain used relies on predicting grass growth on Irish dairy farms, during Europe-wide drought and forage crisis of 2018. --- ### From Events to Clarity: The Event-Guided Diffusion Framework for Dehazing Authors: Ling Wang, Yunfan Lu, Wenzong Ma, Huizai Yao, Pengteng Li, Hui Xiong Published: 2025-11-14 Categories: cs.CV arXiv: https://arxiv.org/abs/2511.11944v1 Clear imaging under hazy conditions is a critical task. Prior-based and neural methods have improved results. However, they operate on RGB frames, which suffer from limited dynamic range. Therefore, dehazing remains ill-posed and can erase structure and illumination details. To address this, we use event cameras for dehazing for the \textbf{first time}. Event cameras offer much higher HDR ($120 dBvs.60 dB$) and microsecond latency, therefore they suit hazy scenes. In practice, transferring HDR cues from events to frames is hard because real paired data are scarce. To tackle this, we propose an event-guided diffusion model that utilizes the strong generative priors of diffusion models to reconstruct clear images from hazy inputs by effectively transferring HDR information from events. Specifically, we design an event-guided module that maps sparse HDR event features, \textit{e.g.,} edges, corners, into the diffusion latent space. This clear conditioning provides precise structural guidance during generation, improves visual realism, and reduces semantic drift. For real-world evaluation, we collect a drone dataset in heavy haze (AQI = 341) with synchronized RGB and event sensors. Experiments on two benchmarks and our dataset achieve state-of-the-art results. --- ### Learning the relative composition of EEG signals using pairwise relative shift pretraining Authors: Christopher Sandino, Sayeri Lala, Geeling Chau, Melika Ayoughi, Behrooz Mahasseni, Ellen Zippi, Ali Moin, Erdrin Azemi, Hanlin Goh Published: 2025-11-14 Categories: cs.LG, eess.SP arXiv: https://arxiv.org/abs/2511.11940v1 Self-supervised learning (SSL) offers a promising approach for learning electroencephalography (EEG) representations from unlabeled data, reducing the need for expensive annotations for clinical applications like sleep staging and seizure detection. While current EEG SSL methods predominantly use masked reconstruction strategies like masked autoencoders (MAE) that capture local temporal patterns, position prediction pretraining remains underexplored despite its potential to learn long-range dependencies in neural signals. We introduce PAirwise Relative Shift or PARS pretraining, a novel pretext task that predicts relative temporal shifts between randomly sampled EEG window pairs. Unlike reconstruction-based methods that focus on local pattern recovery, PARS encourages encoders to capture relative temporal composition and long-range dependencies inherent in neural signals. Through comprehensive evaluation on various EEG decoding tasks, we demonstrate that PARS-pretrained transformers consistently outperform existing pretraining strategies in label-efficient and transfer learning settings, establishing a new paradigm for self-supervised EEG representation learning. --- ### Improving Neutrino Oscillation Measurements through Event Classification Authors: Sebastian A. R. Ellis, Daniel C. Hackett, Shirley Weishi Li, Pedro A. N. Machado, Karla Tame-Narvaez Published: 2025-11-14 Categories: hep-ph, cs.AI, cs.LG, hep-ex arXiv: https://arxiv.org/abs/2511.11938v2 Precise neutrino energy reconstruction is essential for next-generation long-baseline oscillation experiments, yet current methods remain limited by large uncertainties in neutrino-nucleus interaction modeling. Even so, it is well established that different interaction channels produce systematically varying amounts of missing energy and therefore yield different reconstruction performance--information that standard calorimetric approaches do not exploit. We introduce a strategy that incorporates this structure by classifying events according to their underlying interaction type prior to energy reconstruction. Using supervised machine-learning techniques trained on labeled generator events, we leverage intrinsic kinematic differences among quasi-elastic scattering, meson-exchange current, resonance production, and deep-inelastic scattering processes. A cross-generator testing framework demonstrates that this classification approach is robust to microphysics mismodeling and, when applied to a simulated DUNE $ν_μ$ disappearance analysis, yields improved accuracy and sensitivity at the 10-20% level. These results highlight a practical path toward reducing reconstruction-driven systematics in future oscillation measurements. --- ### A Deep Learning Framework for Thyroid Nodule Segmentation and Malignancy Classification from Ultrasound Images Authors: Omar Abdelrazik, Mohamed Elsayed, Noorul Wahab, Nasir Rajpoot, Adam Shephard Published: 2025-11-14 Categories: eess.IV, cs.AI, cs.CV, cs.LG arXiv: https://arxiv.org/abs/2511.11937v1 Ultrasound-based risk stratification of thyroid nodules is a critical clinical task, but it suffers from high inter-observer variability. While many deep learning (DL) models function as "black boxes," we propose a fully automated, two-stage framework for interpretable malignancy prediction. Our method achieves interpretability by forcing the model to focus only on clinically relevant regions. First, a TransUNet model automatically segments the thyroid nodule. The resulting mask is then used to create a region of interest around the nodule, and this localised image is fed directly into a ResNet-18 classifier. We evaluated our framework using 5-fold cross-validation on a clinical dataset of 349 images, where it achieved a high F1-score of 0.852 for predicting malignancy. To validate its performance, we compared it against a strong baseline using a Random Forest classifier with hand-crafted morphological features, which achieved an F1-score of 0.829. The superior performance of our DL framework suggests that the implicit visual features learned from the localised nodule are more predictive than explicit shape features alone. This is the first fully automated end-to-end pipeline for both detecting thyroid nodules on ultrasound images and predicting their malignancy. --- ### SurvBench: A Standardised Preprocessing Pipeline for Multi-Modal Electronic Health Record Survival Analysis Authors: Munib Mesinovic, Tingting Zhu Published: 2025-11-14 Categories: cs.LG arXiv: https://arxiv.org/abs/2511.11935v2 Deep-learning survival models for electronic health record (EHR) data are hard to compare across papers because the upstream preprocessing step, which includes cohort definition, time discretisation, missingness handling, and censoring rules, is typically undocumented and inconsistent. A reported difference in concordance between two mortality models can therefore reflect any of these choices rather than a modelling contribution. We present SurvBench, an open-source preprocessing pipeline that converts raw PhysioNet exports into model-ready tensors for survival analysis. SurvBench covers four critical-care databases (MIMIC-IV, eICU, MC-MED, HiRID) and four input modalities: time-series vitals and laboratory values, static demographics, International Classification of Diseases (ICD) codes, and radiology report embeddings. Every preprocessing decision is controlled through YAML configuration. Imputation, scaling, and feature filtering are fit on the training fold only. Missingness is recorded as a binary mask alongside each feature tensor. The pipeline handles single-risk endpoints (in-hospital and in-ICU mortality) and competing-risks endpoints (a three-way emergency-department admission pathway, with home discharge treated as administrative censoring). We also provide support for harmonised cross-dataset external validation between eICU and MIMIC-IV. SurvBench is publicly available at https://github.com/munibmesinovic/SurvBench, providing a robust platform that future deep-learning EHR survival work, especially nascent multi-modal approaches, can be measured against under matched preprocessing. --- ### A Systematic Analysis of Out-of-Distribution Detection Under Representation and Training Paradigm Shifts Authors: Claudio César Claros Olivares, Austin J. Brockmeier Published: 2025-11-14 Categories: cs.LG, cs.CV arXiv: https://arxiv.org/abs/2511.11934v3 We present a systematic benchmark of out-of-distribution (OOD) detection CSFs through a representation-centric lens. Our study spans CNN and ViT backbones, multiple training paradigms, four image-classification source datasets (CIFAR-10, CIFAR-100, SuperCIFAR-100, and TinyImageNet), and OOD datasets grouped into near, mid, and far regimes using CLIP-derived semantic distances. To compare CSFs across these settings, we employ a multiple-comparison-controlled rank pipeline that identifies top cliques of statistically indistinguishable winners under threshold-free ranking metrics (AURC and AUGRC). The main empirical finding is that the competitive detector family depends more on the learned representation than on score design alone. For both CNNs and ViTs, simple probabilistic scores dominate misclassification detection. On CNNs, margin-based scores are strongest in near-OOD regimes, while geometry-aware scores such as NNGuide, fDBD, and CTM become more competitive as shift severity increases. On fine-tuned ViTs, the top cliques are led mainly by reconstruction- and residual-based scores. To interpret these ranking shifts, we analyze the last-layer representation using Neural Collapse (NC) metrics. The resulting picture is consistent across architectures: prototype- and boundary-aware scores become stronger when the representation is more collapsed and better aligned with classifier weights, whereas weaker-collapse regimes favor gradient- and manifold-based scores. Building on these insights, we propose two contributions: a simple PCA-based projection-filtering procedure that improves detector performance, and an approach that uses NC measurements computed from a trained classifier to predict its competitive out-of-distribution detector shortlist, without requiring any additional OOD data. --- ### InData: Towards Secure Multi-Step, Tool-Based Data Analysis Authors: Karthikeyan K, Raghuveer Thirukovalluru, Bhuwan Dhingra, David Edwin Carlson Published: 2025-11-14 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2511.11933v1 Large language model agents for data analysis typically generate and execute code directly on databases. However, when applied to sensitive data, this approach poses significant security risks. To address this issue, we propose a security-motivated alternative: restrict LLMs from direct code generation and data access, and require them to interact with data exclusively through a predefined set of secure, verified tools. Although recent tool-use benchmarks exist, they primarily target tool selection and simple execution rather than the compositional, multi-step reasoning needed for complex data analysis. To reduce this gap, we introduce Indirect Data Engagement (InData), a dataset designed to assess LLMs' multi-step tool-based reasoning ability. InData includes data analysis questions at three difficulty levels--Easy, Medium, and Hard--capturing increasing reasoning complexity. We benchmark 15 open-source LLMs on InData and find that while large models (e.g., gpt-oss-120b) achieve high accuracy on Easy tasks (97.3%), performance drops sharply on Hard tasks (69.6%). These results show that current LLMs still lack robust multi-step tool-based reasoning ability. With InData, we take a step toward enabling the development and evaluation of LLMs with stronger multi-step tool-use capabilities. We will publicly release the dataset and code. --- ### Enhancing XR Auditory Realism via Multimodal Scene-Aware Acoustic Rendering Authors: Tianyu Xu, Jihan Li, Penghe Zu, Pranav Sahay, Maruchi Kim, Jack Obeng-Marnu, Farley Miller, Xun Qian, Katrina Passarella, Mahitha Rachumalla, Rajeev Nongpiur, D. Shin Published: 2025-11-14 Categories: cs.HC, cs.CV, cs.LG, cs.SD arXiv: https://arxiv.org/abs/2511.11930v1 In Extended Reality (XR), rendering sound that accurately simulates real-world acoustics is pivotal in creating lifelike and believable virtual experiences. However, existing XR spatial audio rendering methods often struggle with real-time adaptation to diverse physical scenes, causing a sensory mismatch between visual and auditory cues that disrupts user immersion. To address this, we introduce SAMOSA, a novel on-device system that renders spatially accurate sound by dynamically adapting to its physical environment. SAMOSA leverages a synergistic multimodal scene representation by fusing real-time estimations of room geometry, surface materials, and semantic-driven acoustic context. This rich representation then enables efficient acoustic calibration via scene priors, allowing the system to synthesize a highly realistic Room Impulse Response (RIR). We validate our system through technical evaluation using acoustic metrics for RIR synthesis across various room configurations and sound types, alongside an expert evaluation (N=12). Evaluation results demonstrate SAMOSA's feasibility and efficacy in enhancing XR auditory realism. --- ### Beyond the Laplacian: Interpolated Spectral Augmentation for Graph Neural Networks Authors: Ziyao Cui, Edric Tam Published: 2025-11-14 Categories: cs.LG arXiv: https://arxiv.org/abs/2511.11928v2 Graph neural networks (GNNs) are fundamental tools in graph machine learning. The performance of GNNs relies crucially on the availability of informative node features, which can be limited or absent in real-life datasets and applications. A natural remedy is to augment the node features with embeddings computed from eigenvectors of the graph Laplacian matrix. While it is natural to default to Laplacian spectral embeddings, which capture meaningful graph connectivity information, we ask whether spectral embeddings from alternative graph matrices can also provide useful representations for learning. We introduce Interpolated Laplacian Embeddings (ILEs), which are derived from a simple yet expressive family of graph matrices. Using tools from spectral graph theory, we offer a straightforward interpretation of the structural information that ILEs capture. We demonstrate through simulations and experiments on real-world datasets that feature augmentation via ILEs can improve performance across commonly used GNN architectures. Our work offers a straightforward and practical approach that broadens the practitioner's spectral augmentation toolkit when node features are limited. --- ### Sparse corruption in low-rank matrix inference: the PCA benchmark Authors: Urte Adomaityte, Gabriele Sicuro, Pierpaolo Vivo Published: 2025-11-14 Categories: stat.ML, cond-mat.dis-nn, cond-mat.stat-mech, cs.LG arXiv: https://arxiv.org/abs/2511.11927v2 Principal Component Analysis (PCA) is a standard tool for extracting a low-rank signal from noisy observations. It is known that applying PCA to a rank-one signal corrupted by a dense, homogeneous noise, in the large matrix size limit, the celebrated BBP transition occurs, where the emergence of an outlying eigenvalue and the alignment of the corresponding eigenvector occur at the same critical signal strength. Here we study the case of sparse noise corruption. The noise matrix is modelled as the adjacency matrix of a weighted undirected graph with finite average connectivity. Using the replica method, we analytically compute the typical top eigenvalue, the top eigenvector component density, and the squared overlap with the signal, through recursive distributional equations solved by population dynamics. We identify two signal-strength transitions as functions of graph connectivity: $θ_{\rm crit}$, marking signal recovery by the top eigenvector and generalising the BBP transition, and $θ_{\rm b}$, where the signal-related eigenvalue detaches from the bulk. For noise with nonzero mean, these transitions need not coincide because of a structural sparse-graph outlier, leading to a discontinuous transition in the squared overlap with the top eigenvector. The same top-eigenpair formalism also predicts the overlap of the signal with the eigenvector associated with the second largest eigenvalue when the signal eigenvalue is an outlier but remains below the structural outlier, where the transition is continuous. We specialise the equations to Poissonian and Random Regular degree distributions, recover dense-noise results in the large-connectivity limit, and validate the theory by numerical diagonalisation of large matrices. --- ### A Neuromorphic Architecture for Scalable Event-Based Control Authors: Yongkang Huo, Fulvio Forni, Rodolphe Sepulchre Published: 2025-11-14 Categories: cs.AI arXiv: https://arxiv.org/abs/2511.11924v2 This paper introduces the ``rebound Winner-Take-All (RWTA)" motif as the basic element of a scalable neuromorphic control architecture. From the cellular level to the system level, the resulting architecture combines the reliability of discrete computation and the tunability of continuous regulation: it inherits the discrete computation capabilities of winner-take-all state machines and the continuous tuning capabilities of excitable biophysical circuits. The proposed event-based framework addresses continuous rhythmic generation and discrete decision-making in a unified physical modeling language. We illustrate the versatility, robustness, and modularity of the architecture through the nervous system design of a snake robot. --- ### Additive Large Language Models for Semi-Structured Text Authors: Karthikeyan K, Raghuveer Thirukovalluru, David Carlson Published: 2025-11-14 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2511.11922v1 Large Language Models have advanced clinical text classification, but their opaque predictions remain a critical barrier to practical adoption in research and clinical settings where investigators and physicians need to understand which parts of a patient's record drive risk signals. To address this challenge, we introduce \textbf{CALM}, short for \textbf{Classification with Additive Large Language Models}, an interpretable framework for semi-structured text where inputs are composed of semantically meaningful components, such as sections of an admission note or question-answer fields from an intake form. CALM predicts outcomes as the additive sum of each component's contribution, making these contributions part of the forward computation itself and enabling faithful explanations at both the patient and population level. The additive structure also enables clear visualizations, such as component-level risk curves similar to those used in generalized additive models, making the learned relationships easier to inspect and communicate. Although CALM expects semi-structured inputs, many clinical documents already have this form, and similar structure can often be automatically extracted from free-text notes. CALM achieves performance comparable to conventional LLM classifiers while improving trust, supporting quality-assurance checks, and revealing clinically meaningful patterns during model development and auditing. --- ### Looking Forward: Challenges and Opportunities in Agentic AI Reliability Authors: Liudong Xing, Janet, Lin Published: 2025-11-14 Categories: cs.AI, cs.ET, cs.PF arXiv: https://arxiv.org/abs/2511.11921v1 This chapter presents perspectives for challenges and future development in building reliable AI systems, particularly, agentic AI systems. Several open research problems related to mitigating the risks of cascading failures are discussed. The chapter also sheds lights on research challenges and opportunities in aspects including dynamic environments, inconsistent task execution, unpredictable emergent behaviors, as well as resource-intensive reliability mechanisms. In addition, several research directions along the line of testing and evaluating reliability of agentic AI systems are also discussed. --- ### Batch Matrix-form Equations and Implementation of Multilayer Perceptrons Authors: Wieger Wesselink, Bram Grooten, Huub van de Wetering, Qiao Xiao, Decebal Constantin Mocanu Published: 2025-11-14 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2511.11918v1 Multilayer perceptrons (MLPs) remain fundamental to modern deep learning, yet their algorithmic details are rarely presented in complete, explicit \emph{batch matrix-form}. Rather, most references express gradients per sample or rely on automatic differentiation. Although automatic differentiation can achieve equally high computational efficiency, the usage of batch matrix-form makes the computational structure explicit, which is essential for transparent, systematic analysis, and optimization in settings such as sparse neural networks. This paper fills that gap by providing a mathematically rigorous and implementation-ready specification of MLPs in batch matrix-form. We derive forward and backward equations for all standard and advanced layers, including batch normalization and softmax, and validate all equations using the symbolic mathematics library SymPy. From these specifications, we construct uniform reference implementations in NumPy, PyTorch, JAX, TensorFlow, and a high-performance C++ backend optimized for sparse operations. Our main contributions are: (1) a complete derivation of batch matrix-form backpropagation for MLPs, (2) symbolic validation of all gradient equations, (3) uniform Python and C++ reference implementations grounded in a small set of matrix primitives, and (4) demonstration of how explicit formulations enable efficient sparse computation. Together, these results establish a validated, extensible foundation for understanding, teaching, and researching neural network algorithms. --- ### An Analysis of Architectural Impact on LLM-based Abstract Visual Reasoning: A Systematic Benchmark on RAVEN-FAIR Authors: Sinan Urgun, Seçkin Arı Published: 2025-11-14 Categories: cs.AI arXiv: https://arxiv.org/abs/2511.11916v1 This study aims to systematically evaluate the performance of large language models (LLMs) in abstract visual reasoning problems. We examined four LLM models (GPT-4.1-Mini, Claude-3.5-Haiku, Gemini-1.5-Flash, Llama-3.3-70b) utilizing four different reasoning architectures (single-shot, embedding-controlled repetition, self-reflection, and multi-agent) on the RAVEN-FAIR dataset. Visual responses generated through a three-stage process (JSON extraction, LLM reasoning, and Tool Function) were evaluated using SSIM and LPIPS metrics; Chain-of-Thought scores and error types (semantic hallucination, numeric misperception) were analyzed. Results demonstrate that GPT-4.1-Mini consistently achieved the highest overall accuracy across all architectures, indicating a strong reasoning capability. While the multi-agent architecture occasionally altered semantic and numeric balance across models, these effects were not uniformly beneficial. Instead, each model exhibited distinct sensitivity patterns to architectural design, underscoring that reasoning effectiveness remains model-specific. Variations in response coverage further emerged as a confounding factor that complicates direct cross-architecture comparison. To estimate the upper-bound performance of each configuration, we report the best of five independent runs, representing a best-case scenario rather than an averaged outcome. This multi-run strategy aligns with recent recommendations, which emphasize that single-run evaluations are fragile and may lead to unreliable conclusions. --- ### Forgetting-MarI: LLM Unlearning via Marginal Information Regularization Authors: Shizhou Xu, Yuan Ni, Stefan Broecker, Thomas Strohmer Published: 2025-11-14 Categories: cs.AI, cs.CL, cs.CR, cs.IT, cs.LG arXiv: https://arxiv.org/abs/2511.11914v3 As AI models are trained on ever-expanding datasets, the ability to remove the influence of specific data from trained models has become essential for privacy protection and regulatory compliance. Unlearning addresses this challenge by selectively removing parametric knowledge from the trained models without retraining from scratch, which is critical for resource-intensive models such as Large Language Models (LLMs). Existing unlearning methods often degrade model performance by removing more information than necessary when attempting to ''forget'' specific data. We introduce Forgetting-MarI, an LLM unlearning framework that provably removes only the additional (marginal) information contributed by the data to be unlearned, while preserving the information supported by the data to be retained. By penalizing marginal information, our method yields an explicit upper bound on the unlearn dataset's residual influence in the trained models, providing provable undetectability. Extensive experiments confirm that our approach outperforms current state-of-the-art unlearning methods, delivering reliable forgetting and better preserved general model performance across diverse benchmarks. This advancement represents an important step toward making AI systems more controllable and compliant with privacy and copyright regulations without compromising their effectiveness. --- ### A Systematic Study of Model Extraction Attacks on Graph Foundation Models Authors: Haoyan Xu, Ruizhi Qian, Jiate Li, Yushun Dong, Minghao Lin, Hanson Yan, Zhengtao Yao, Qinghua Liu, Junhao Dong, Ruopeng Huang, Yue Zhao, Mengyuan Li Published: 2025-11-14 Categories: cs.LG, cs.CR arXiv: https://arxiv.org/abs/2511.11912v1 Graph machine learning has advanced rapidly in tasks such as link prediction, anomaly detection, and node classification. As models scale up, pretrained graph models have become valuable intellectual assets because they encode extensive computation and domain expertise. Building on these advances, Graph Foundation Models (GFMs) mark a major step forward by jointly pretraining graph and text encoders on massive and diverse data. This unifies structural and semantic understanding, enables zero-shot inference, and supports applications such as fraud detection and biomedical analysis. However, the high pretraining cost and broad cross-domain knowledge in GFMs also make them attractive targets for model extraction attacks (MEAs). Prior work has focused only on small graph neural networks trained on a single graph, leaving the security implications for large-scale and multimodal GFMs largely unexplored. This paper presents the first systematic study of MEAs against GFMs. We formalize a black-box threat model and define six practical attack scenarios covering domain-level and graph-specific extraction goals, architectural mismatch, limited query budgets, partial node access, and training data discrepancies. To instantiate these attacks, we introduce a lightweight extraction method that trains an attacker encoder using supervised regression of graph embeddings. Even without contrastive pretraining data, this method learns an encoder that stays aligned with the victim text encoder and preserves its zero-shot inference ability on unseen graphs. Experiments on seven datasets show that the attacker can approximate the victim model using only a tiny fraction of its original training cost, with almost no loss in accuracy. These findings reveal that GFMs greatly expand the MEA surface and highlight the need for deployment-aware security defenses in large-scale graph learning systems. --- ### Seeing the Forest and the Trees: Query-Aware Tokenizer for Long-Video Multimodal Language Models Authors: Siyou Li, Huanan Wu, Juexi Shao, Yinghao Ma, Yujian Gan, Yihao Luo, Yuwei Wang, Dong Nie, Lu Wang, Wenqing Wu, Le Zhang, Massimo Poesio et al. Published: 2025-11-14 Categories: cs.CV arXiv: https://arxiv.org/abs/2511.11910v3 Despite the recent advances in the video understanding ability of multimodal large language models (MLLMs), long video understanding remains a challenge. One of the main issues is that the number of vision tokens grows linearly with video length, which causes an explosion in attention cost, memory, and latency. To solve this challenge, we present Query-aware Token Selector (\textbf{QTSplus}), a lightweight yet powerful visual token selection module that serves as an information gate between the vision encoder and LLMs. Given a text query and video tokens, QTSplus dynamically selects the most important visual evidence for the input text query by (i) scoring visual tokens via cross-attention, (ii) \emph{predicting} an instance-specific retention budget based on the complexity of the query, and (iii) \emph{selecting} Top-$n$ tokens with a differentiable straight-through estimator during training and a hard gate at inference. Furthermore, a small re-encoder preserves temporal order using absolute time information, enabling second-level localization while maintaining global coverage. Integrated into Qwen2.5-VL, QTSplus compresses the vision stream by up to \textbf{89\%} and reduces end-to-end latency by \textbf{28\%} on long videos. The evaluation on eight long video understanding benchmarks shows near-parity accuracy overall when compared with the original Qwen models and outperforms the original model by \textbf{+20.5} and \textbf{+5.6} points respectively on TempCompass direction and order accuracies. These results show that QTSplus is an effective, general mechanism for scaling MLLMs to real-world long-video scenarios while preserving task-relevant evidence. --- ### PI-NAIM: Path-Integrated Neural Adaptive Imputation Model Authors: Afifa Khaled, Ebrahim Hamid Sumiea Published: 2025-11-14 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2511.11908v1 Medical imaging and multi-modal clinical settings often face the challange of missing modality in their diagnostic pipelines. Existing imputation methods either lack representational capacity or are computationally expensive. We propose PI-NAIM, a novel dual-path architecture that dynamically routes samples to optimized imputation approaches based on missingness complexity. Our framework integrates: (1) intelligent path routing that directs low missingness samples to efficient statistical imputation (MICE) and complex patterns to powerful neural networks (GAIN with temporal analysis); (2) cross-path attention fusion that leverages missingness-aware embeddings to intelligently combine both branches; and (3) end-to-end joint optimization of imputation accuracy and downstream task performance. Extensive experiments on MIMIC-III and multimodal benchmarks demonstrate state-of-the-art performance, achieving RMSE of 0.108 (vs. baselines' 0.119-0.152) and substantial gains in downstream tasks with an AUROC of 0.812 for mortality prediction. PI-NAIM's modular design enables seamless integration into vision pipelines handling incomplete sensor measurements, missing modalities, or corrupted inputs, providing a unified solution for real-world scenario. The code is publicly available at https://github.com/AfifaKhaled/PI-NAIM-Path-Integrated-Neural-Adaptive-Imputation-Model --- ### KVSwap: Disk-aware KV Cache Offloading for Long-Context On-device Inference Authors: Huawei Zhang, Chunwei Xia, Zheng Wang Published: 2025-11-14 Categories: cs.DC, cs.AI arXiv: https://arxiv.org/abs/2511.11907v2 Language models (LMs) underpin emerging mobile and embedded AI applications like meeting and video summarization and document analysis, which often require processing multiple long-context inputs. Running an LM locally on-device improves privacy, enables offline use, and reduces cost, but long-context inference quickly hits a \emph{memory capacity wall} as the key-value (KV) cache grows linearly with context length and batch size. Existing KV-cache offloading schemes are designed to transfer cache data from GPU memory to CPU memory; however, they are not suitable for embedded and mobile systems, where the CPU and GPU (or NPU) typically share a unified memory and the non-volatile secondary storage (disk) offers limited I/O bandwidth. We present KVSwap, a software framework tailored for local devices that achieves high memory efficiency while effectively leveraging disk storage. KVSwap stores the full cache on disk, uses highly compact in-memory metadata to predict which entries to preload, overlaps computation with hardware-aware disk access, and orchestrates read patterns to match storage device characteristics. Our evaluation shows that across representative LMs and storage types, KVSwap delivers higher throughput under tight memory budgets while maintaining generation quality over existing KV cache offloading schemes. --- ### Robust Bidirectional Associative Memory via Regularization Inspired by the Subspace Rotation Algorithm Authors: Ci Lin, Tet Yeap, Iluju Kiringa Published: 2025-11-14 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2511.11902v2 Bidirectional Associative Memory (BAM) trained with Bidirectional Backpropagation (B-BP) often suffers from poor robustness and high sensitivity to noise and adversarial attacks. To address these issues, we propose a novel gradient-free training algorithm, the Bidirectional Subspace Rotation Algorithm (B-SRA), which significantly improves the robustness and convergence behavior of BAM. Through comprehensive experiments, we identify two key principles -- orthogonal weight matrices (OWM) and gradient-pattern alignment (GPA) -- as central to enhancing the robustness of BAM. Motivated by these findings, we introduce new regularization strategies into B-BP, resulting in models with greatly improved resistance to corruption and adversarial perturbations. We further conduct an ablation study across different training strategies to determine the most robust configuration and evaluate BAM's performance under a variety of attack scenarios and memory capacities, including 50, 100, and 200 associative pairs. Among all methods, the SAME configuration, which integrates both OWM and GPA, achieves the strongest resilience. Overall, our results demonstrate that B-SRA and the proposed regularization strategies lead to substantially more robust associative memories and open new directions for building resilient neural architectures. --- ### End to End AI System for Surgical Gesture Sequence Recognition and Clinical Outcome Prediction Authors: Xi Li, Nicholas Matsumoto, Ujjwal Pasupulety, Atharva Deo, Cherine Yang, Jay Moran, Miguel E. Hernandez, Peter Wager, Jasmine Lin, Jeanine Kim, Alvin C. Goh, Christian Wagner et al. Published: 2025-11-14 Categories: cs.AI, cs.CV arXiv: https://arxiv.org/abs/2511.11899v1 Fine-grained analysis of intraoperative behavior and its impact on patient outcomes remain a longstanding challenge. We present Frame-to-Outcome (F2O), an end-to-end system that translates tissue dissection videos into gesture sequences and uncovers patterns associated with postoperative outcomes. Leveraging transformer-based spatial and temporal modeling and frame-wise classification, F2O robustly detects consecutive short (~2 seconds) gestures in the nerve-sparing step of robot-assisted radical prostatectomy (AUC: 0.80 frame-level; 0.81 video-level). F2O-derived features (gesture frequency, duration, and transitions) predicted postoperative outcomes with accuracy comparable to human annotations (0.79 vs. 0.75; overlapping 95% CI). Across 25 shared features, effect size directions were concordant with small differences (~ 0.07), and strong correlation (r = 0.96, p < 1e-14). F2O also captured key patterns linked to erectile function recovery, including prolonged tissue peeling and reduced energy use. By enabling automatic interpretable assessment, F2O establishes a foundation for data-driven surgical feedback and prospective clinical decision support. --- ### Prompt Triage: Structured Optimization Enhances Vision-Language Model Performance on Medical Imaging Benchmarks Authors: Arnav Singhvi, Vasiliki Bikia, Asad Aali, Akshay Chaudhari, Roxana Daneshjou Published: 2025-11-14 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2511.11898v1 Vision-language foundation models (VLMs) show promise for diverse imaging tasks but often underperform on medical benchmarks. Prior efforts to improve performance include model finetuning, which requires large domain-specific datasets and significant compute, or manual prompt engineering, which is hard to generalize and often inaccessible to medical institutions seeking to deploy these tools. These challenges motivate interest in approaches that draw on a model's embedded knowledge while abstracting away dependence on human-designed prompts to enable scalable, weight-agnostic performance improvements. To explore this, we adapt the Declarative Self-improving Python (DSPy) framework for structured automated prompt optimization in medical vision-language systems through a comprehensive, formal evaluation. We implement prompting pipelines for five medical imaging tasks across radiology, gastroenterology, and dermatology, evaluating 10 open-source VLMs with four prompt optimization techniques. Optimized pipelines achieved a median relative improvement of 53% over zero-shot prompting baselines, with the largest gains ranging from 300% to 3,400% on tasks where zero-shot performance is low. These results highlight the substantial potential of applying automated prompt optimization to medical AI systems, demonstrating significant gains for vision-based applications requiring accurate clinical image interpretation. By reducing dependence on prompt design to elicit intended outputs, these techniques allow clinicians to focus on patient care and clinical decision-making. Furthermore, our experiments offer scalability and preserve data privacy, demonstrating performance improvement on open-source VLMs. We publicly release our evaluation pipelines to support reproducible research on specialized medical tasks, available at https://github.com/DaneshjouLab/prompt-triage-lab. --- ### VULPO: Context-Aware Vulnerability Detection via On-Policy LLM Optimization Authors: Youpeng Li, Fuxun Yu, Weiliang Qi, Xinda Wang Published: 2025-11-14 Categories: cs.CR, cs.AI, cs.SE arXiv: https://arxiv.org/abs/2511.11896v3 Large language models (LLMs) have recently shown strong potential in vulnerability detection (VD). However, accurately detecting vulnerabilities in real-world repositories requires reasoning over complex contextual interactions. Existing LLM-based VD approaches remain limited because current datasets lack complete contextual information and high-quality reasoning supervision, while existing optimization methods primarily rely on coarse outcome-centric supervision signals that fail to model the vulnerability reasoning process. To address these limitations, we first construct ContextVul, a new dataset that augments high-quality function-level vulnerability benchmarks with repository-level contextual information and curated vulnerability reasoning traces. Building upon ContextVul, we introduce a two-stage optimization framework consisting of lightweight cold-start supervised fine-tuning followed by vulnerability-adaptive on-policy optimization (VULPO). VULPO incorporates multidimensional rewards that jointly evaluate vulnerability identification, vulnerability-relevant localization, and causal reasoning quality, along with difficulty-adaptive reward scaling to mitigate reward hacking and improve RL effectiveness. Extensive experiments demonstrate the superiority of VULPO for context-aware VD. Our VULPO-4B, the first specialized vulnerability reasoning LLM, substantially outperforms existing VD baselines, improving Pairwise Pass@1 by 203% relative to Qwen3-4B and achieving competitive performance against a 150% larger-scale LLM, DeepSeek-V3.1. --- ### Chain-of-Generation: Progressive Latent Diffusion for Text-Guided Molecular Design Authors: Lingxiao Li, Haobo Zhang, Bin Chen, Jiayu Zhou Published: 2025-11-14 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2511.11894v1 Text-conditioned molecular generation aims to translate natural-language descriptions into chemical structures, enabling scientists to specify functional groups, scaffolds, and physicochemical constraints without handcrafted rules. Diffusion-based models, particularly latent diffusion models (LDMs), have recently shown promise by performing stochastic search in a continuous latent space that compactly captures molecular semantics. Yet existing methods rely on one-shot conditioning, where the entire prompt is encoded once and applied throughout diffusion, making it hard to satisfy all the requirements in the prompt. We discuss three outstanding challenges of one-shot conditioning generation, including the poor interpretability of the generated components, the failure to generate all substructures, and the overambition in considering all requirements simultaneously. We then propose three principles to address those challenges, motivated by which we propose Chain-of-Generation (CoG), a training-free multi-stage latent diffusion framework. CoG decomposes each prompt into curriculum-ordered semantic segments and progressively incorporates them as intermediate goals, guiding the denoising trajectory toward molecules that satisfy increasingly rich linguistic constraints. To reinforce semantic guidance, we further introduce a post-alignment learning phase that strengthens the correspondence between textual and molecular latent spaces. Extensive experiments on benchmark and real-world tasks demonstrate that CoG yields higher semantic alignment, diversity, and controllability than one-shot baselines, producing molecules that more faithfully reflect complex, compositional prompts while offering transparent insight into the generation process. --- ### FLEX: Feature Importance from Layered Counterfactual Explanations Authors: Nawid Keshtmand, Roussel Desmond Nzoyem, Jeffrey Nicholas Clark Published: 2025-11-14 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2511.11891v1 Machine learning models achieve state-of-the-art performance across domains, yet their lack of interpretability limits safe deployment in high-stakes settings. Counterfactual explanations are widely used to provide actionable "what-if" recourse, but they typically remain instance-specific and do not quantify which features systematically drive outcome changes within coherent regions of the feature space or across an entire dataset. We introduce FLEX (Feature importance from Layered counterfactual EXplanations), a model- and domain-agnostic framework that converts sets of counterfactuals into feature change frequency scores at local, regional, and global levels. FLEX generalises local change-frequency measures by aggregating across instances and neighbourhoods, offering interpretable rankings that reflect how often each feature must change to flip predictions. The framework is compatible with different counterfactual generation methods, allowing users to emphasise characteristics such as sparsity, feasibility, or actionability, thereby tailoring the derived feature importances to practical constraints. We evaluate FLEX on two contrasting tabular tasks: traffic accident severity prediction and loan approval, and compare FLEX to SHAP- and LIME-derived feature importance values. Results show that (i) FLEX's global rankings correlate with SHAP while surfacing additional drivers, and (ii) regional analyses reveal context-specific factors that global summaries miss. FLEX thus bridges the gap between local recourse and global attribution, supporting transparent and intervention-oriented decision-making in risk-sensitive applications. --- ### Advancing Annotat3D with Harpia: A CUDA-Accelerated Library For Large-Scale Volumetric Data Segmentation Authors: Camila Machado de Araujo, Egon P. B. S. Borges, Ricardo Marcelo Canteiro Grangeiro, Allan Pinto Published: 2025-11-14 Categories: cs.CV, cs.DC arXiv: https://arxiv.org/abs/2511.11890v1 High-resolution volumetric imaging techniques, such as X-ray tomography and advanced microscopy, generate increasingly large datasets that challenge existing tools for efficient processing, segmentation, and interactive exploration. This work introduces new capabilities to Annotat3D through Harpia, a new CUDA-based processing library designed to support scalable, interactive segmentation workflows for large 3D datasets in high-performance computing (HPC) and remote-access environments. Harpia features strict memory control, native chunked execution, and a suite of GPU-accelerated filtering, annotation, and quantification tools, enabling reliable operation on datasets exceeding single-GPU memory capacity. Experimental results demonstrate significant improvements in processing speed, memory efficiency, and scalability compared to widely used frameworks such as NVIDIA cuCIM and scikit-image. The system's interactive, human-in-the-loop interface, combined with efficient GPU resource management, makes it particularly suitable for collaborative scientific imaging workflows in shared HPC infrastructures. --- ### Flash-Fusion: Enabling Expressive, Low-Latency Queries on IoT Sensor Streams with LLMs Authors: Kausar Patherya, Ashutosh Dhekne, Francisco Romero Published: 2025-11-14 Categories: cs.DC, cs.AI, cs.DB arXiv: https://arxiv.org/abs/2511.11885v1 Smart cities and pervasive IoT deployments have generated interest in IoT data analysis across transportation and urban planning. At the same time, Large Language Models offer a new interface for exploring IoT data - particularly through natural language. Users today face two key challenges when working with IoT data using LLMs: (1) data collection infrastructure is expensive, producing terabytes of low-level sensor readings that are too granular for direct use, and (2) data analysis is slow, requiring iterative effort and technical expertise. Directly feeding all IoT telemetry to LLMs is impractical due to finite context windows, prohibitive token costs at scale, and non-interactive latencies. What is missing is a system that first parses a user's query to identify the analytical task, then selects the relevant data slices, and finally chooses the right representation before invoking an LLM. We present Flash-Fusion, an end-to-end edge-cloud system that reduces the IoT data collection and analysis burden on users. Two principles guide its design: (1) edge-based statistical summarization (achieving 73.5% data reduction) to address data volume, and (2) cloud-based query planning that clusters behavioral data and assembles context-rich prompts to address data interpretation. We deploy Flash-Fusion on a university bus fleet and evaluate it against a baseline that feeds raw data to a state-of-the-art LLM. Flash-Fusion achieves a 95% latency reduction and 98% decrease in token usage and cost while maintaining high-quality responses. It enables personas across disciplines - safety officers, urban planners, fleet managers, and data scientists - to efficiently iterate over IoT data without the burden of manual query authoring or preprocessing. --- ### Context-Emotion Aware Therapeutic Dialogue Generation: A Multi-component Reinforcement Learning Approach to Language Models for Mental Health Support Authors: Eric Hua Qing Zhang, Julia Ive Published: 2025-11-14 Categories: cs.CL arXiv: https://arxiv.org/abs/2511.11884v2 Mental health disorders impose a substantial global socioeconomic burden. While large language models (LLMs) offer 24/7, non-judgmental interactions to address this gap, pretrained models lack contextual coherence and emotional alignment for appropriate therapeutic dialogue. Existing methods suffer from three critical methodological gaps: 1) Supervised Fine-Tuning (SFT) produces repetitive, context-insensitive outputs that fail to balance clinical accuracy with genuine empathy; 2) Reinforcement Learning (RL)-based therapeutic systems rely on generic reward functions (e.g., BLEU, ROUGE) that prioritise lexical similarity over clinical-specific emotional appropriateness and contextual relevance; 3) LLMs are resource-intensive and pose data privacy risks, making local deployment in clinical settings infeasible. To address these gaps, this study investigates the application of SFT and RL techniques to enhance GPT-2's capacity for therapeutic dialogue generation. The methodology restructured input formats to enable simultaneous processing of contextual information and emotional states alongside user input, employing a novel multi-component reward function that explicitly aligns model outputs with professional therapeutic logic (not just lexical overlap) and annotated emotions. Results demonstrated substantial improvements through RLs over baseline GPT-2 across multiple evaluation metrics: BLEU (0.0111), ROUGE-1 (0.1397), ROUGE-2 (0.0213), ROUGE-L (0.1317), and METEOR (0.0581). LLM evaluation confirmed high contextual relevance and professionalism, while RL achieved 99.34% emotion accuracy compared to 66.96% for baseline GPT-2. These findings demonstrate RL's effectiveness in developing therapeutic dialogue systems that can serve as valuable assistive tools for therapists, while maintaining essential human clinical oversight. --- ### ClinStructor: AI-Powered Structuring of Unstructured Clinical Texts Authors: Karthikeyan K, Raghuveer Thirukovalluru, David Carlson Published: 2025-11-14 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2511.11883v1 Clinical notes contain valuable, context-rich information, but their unstructured format introduces several challenges, including unintended biases (e.g., gender or racial bias), and poor generalization across clinical settings (e.g., models trained on one EHR system may perform poorly on another due to format differences) and poor interpretability. To address these issues, we present ClinStructor, a pipeline that leverages large language models (LLMs) to convert clinical free-text into structured, task-specific question-answer pairs prior to predictive modeling. Our method substantially enhances transparency and controllability and only leads to a modest reduction in predictive performance (a 2-3% drop in AUC), compared to direct fine-tuning, on the ICU mortality prediction task. ClinStructor lays a strong foundation for building reliable, interpretable, and generalizable machine learning models in clinical environments. --- ### Lacking Data? No worries! How synthetic images can alleviate image scarcity in wildlife surveys: a case study with muskox (Ovibos moschatus) Authors: Simon Durand, Samuel Foucher, Alexandre Delplanque, Joëlle Taillon, Jérôme Théau Published: 2025-11-14 Categories: cs.CV arXiv: https://arxiv.org/abs/2511.11882v2 Accurate population estimates are essential for wildlife management, providing critical insights into species abundance and distribution. Traditional survey methods, including visual aerial counts and GNSS telemetry tracking, are widely used to monitor muskox populations in Arctic regions. These approaches are resource intensive and constrained by logistical challenges. Advances in remote sensing, artificial intelligence, and high resolution aerial imagery offer promising alternatives for wildlife detection. Yet, the effectiveness of deep learning object detection models (ODMs) is often limited by small datasets, making it challenging to train robust ODMs for sparsely distributed species like muskoxen. This study investigates the integration of synthetic imagery (SI) to supplement limited training data and improve muskox detection in zero shot (ZS) and few-shot (FS) settings. We compared a baseline model trained on real imagery with 5 ZS and 5 FS models that incorporated progressively more SI in the training set. For the ZS models, where no real images were included in the training set, adding SI improved detection performance. As more SI were added, performance in precision, recall and F1 score increased, but eventually plateaued, suggesting diminishing returns when SI exceeded 100% of the baseline model training dataset. For FS models, combining real and SI led to better recall and slightly higher overall accuracy compared to using real images alone, though these improvements were not statistically significant. Our findings demonstrate the potential of SI to train accurate ODMs when data is scarce, offering important perspectives for wildlife monitoring by enabling rare or inaccessible species to be monitored and to increase monitoring frequency. This approach could be used to initiate ODMs without real data and refine it as real images are acquired over time. --- ### Better LLM Reasoning via Dual-Play Authors: Zhengxin Zhang, Chengyu Huang, Aochong Oliver Li, Claire Cardie Published: 2025-11-14 Categories: cs.LG, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2511.11881v3 Large Language Models (LLMs) have achieved remarkable progress through Reinforcement Learning with Verifiable Rewards (RLVR), yet still rely heavily on external supervision (e.g., curated labels). Adversarial learning, particularly through self-play, offers a promising alternative that enables models to iteratively learn from themselves - thus reducing reliance on external supervision. Dual-play extends adversarial learning by assigning specialized roles to two models and training them against each other, fostering sustained competition and mutual evolution. Despite its promise, adapting dual-play training to LLMs remains limited, largely due to their susceptibility to reward hacking and training instability. In this paper, we introduce PasoDoble, a novel LLM dual-play framework. PasoDoble adversarially trains two models initialized from the same base model: a Proposer, which generates challenging questions with ground-truth answers, and a Solver, which attempts to solve them. We enrich the Proposer with knowledge from a pre-training dataset to ensure the questions' quality and diversity. To avoid reward hacking, the Proposer is rewarded for producing only valid questions that push the Solver's limit, while the Solver is rewarded for solving them correctly, and both are updated jointly. To further enhance training stability, we introduce an optional offline paradigm that decouples Proposer and Solver updates, alternately updating each for several steps while holding the other fixed. Notably, PasoDoble operates without supervision during training. Experimental results show that PasoDoble can improve the reasoning performance of LLMs. Our project page is available at https://hcy123902.github.io/PasoDoble. --- ### Transformers vs. Recurrent Models for Estimating Forest Gross Primary Production Authors: David Montero, Miguel D. Mahecha, Francesco Martinuzzi, César Aybar, Anne Klosterhalfen, Alexander Knohl, Jesús Anaya, Clemens Mosig, Sebastian Wieneke Published: 2025-11-14 Categories: cs.LG, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2511.11880v1 Monitoring the spatiotemporal dynamics of forest CO$_2$ uptake (Gross Primary Production, GPP), remains a central challenge in terrestrial ecosystem research. While Eddy Covariance (EC) towers provide high-frequency estimates, their limited spatial coverage constrains large-scale assessments. Remote sensing offers a scalable alternative, yet most approaches rely on single-sensor spectral indices and statistical models that are often unable to capture the complex temporal dynamics of GPP. Recent advances in deep learning (DL) and data fusion offer new opportunities to better represent the temporal dynamics of vegetation processes, but comparative evaluations of state-of-the-art DL models for multimodal GPP prediction remain scarce. Here, we explore the performance of two representative models for predicting GPP: 1) GPT-2, a transformer architecture, and 2) Long Short-Term Memory (LSTM), a recurrent neural network, using multivariate inputs. Overall, both achieve similar accuracy. But, while LSTM performs better overall, GPT-2 excels during extreme events. Analysis of temporal context length further reveals that LSTM attains similar accuracy using substantially shorter input windows than GPT-2, highlighting an accuracy-efficiency trade-off between the two architectures. Feature importance analysis reveals radiation as the dominant predictor, followed by Sentinel-2, MODIS land surface temperature, and Sentinel-1 contributions. Our results demonstrate how model architecture, context length, and multimodal inputs jointly determine performance in GPP prediction, guiding future developments of DL frameworks for monitoring terrestrial carbon dynamics. --- ### MedPT: A Massive Medical Question Answering Dataset for Brazilian-Portuguese Speakers Authors: Fernanda Bufon Färber, Iago Alves Brito, Julia Soares Dollis, Pedro Schindler Freire Brasil Ribeiro, Rafael Teixeira Sousa, Arlindo Rodrigues Galvão Filho Published: 2025-11-14 Categories: cs.CL arXiv: https://arxiv.org/abs/2511.11878v2 While large language models (LLMs) show transformative potential in healthcare, their development remains focused on high-resource languages. This creates a critical barrier for other languages, as simple translation fails to capture unique clinical and cultural nuances, such as endemic diseases. To address this, we introduce MedPT, the first large-scale, real-world corpus of patient-doctor interactions for the Brazilian Portuguese medical domain. Comprising 384,095 authentic question-answer pairs and covering over 3,200 distinct health-related conditions, the dataset was refined through a rigorous multi-stage curation protocol that employed a hybrid quantitative-qualitative analysis to filter noise and contextually enrich thousands of ambiguous queries, resulting in a corpus of approximately 57 million tokens. We further utilize of LLM-driven annotation to classify queries into seven semantic types to capture user intent. To validate MedPT's utility, we benchmark it in a medical specialty classification task: fine-tuning a 1.7B parameter model achieves an outstanding 94\% F1-score on a 20-class setup. Furthermore, our qualitative error analysis shows misclassifications are not random but reflect genuine clinical ambiguities (e.g., between comorbid conditions), proving the dataset's deep semantic richness. We publicly release MedPT on Hugging Face to support the development of more equitable, accurate, and culturally-aware medical technologies for the Portuguese-speaking world. --- ### Identifying Imaging Follow-Up in Radiology Reports: A Comparative Analysis of Traditional ML and LLM Approaches Authors: Namu Park, Giridhar Kaushik Ramachandran, Kevin Lybarger, Fei Xia, Ozlem Uzuner, Meliha Yetisgen, Martin Gunn Published: 2025-11-14 Categories: cs.CL arXiv: https://arxiv.org/abs/2511.11867v1 Large language models (LLMs) have shown considerable promise in clinical natural language processing, yet few domain-specific datasets exist to rigorously evaluate their performance on radiology tasks. In this work, we introduce an annotated corpus of 6,393 radiology reports from 586 patients, each labeled for follow-up imaging status, to support the development and benchmarking of follow-up adherence detection systems. Using this corpus, we systematically compared traditional machine-learning classifiers, including logistic regression (LR), support vector machines (SVM), Longformer, and a fully fine-tuned Llama3-8B-Instruct, with recent generative LLMs. To evaluate generative LLMs, we tested GPT-4o and the open-source GPT-OSS-20B under two configurations: a baseline (Base) and a task-optimized (Advanced) setting that focused inputs on metadata, recommendation sentences, and their surrounding context. A refined prompt for GPT-OSS-20B further improved reasoning accuracy. Performance was assessed using precision, recall, and F1 scores with 95% confidence intervals estimated via non-parametric bootstrapping. Inter-annotator agreement was high (F1 = 0.846). GPT-4o (Advanced) achieved the best performance (F1 = 0.832), followed closely by GPT-OSS-20B (Advanced; F1 = 0.828). LR and SVM also performed strongly (F1 = 0.776 and 0.775), underscoring that while LLMs approach human-level agreement through prompt optimization, interpretable and resource-efficient models remain valuable baselines. --- ### FocusSDF: Boundary-Aware Learning for Medical Image Segmentation via Signed Distance Supervision Authors: Muzammal Shafique, Nasir Rahim, Jamil Ahmad, Mohammad Siadat, Khalid Malik, Ghaus Malik Published: 2025-11-14 Categories: cs.CV arXiv: https://arxiv.org/abs/2511.11864v1 Segmentation of medical images constitutes an essential component of medical image analysis, providing the foundation for precise diagnosis and efficient therapeutic interventions in clinical practices. Despite substantial progress, most segmentation models do not explicitly encode boundary information; as a result, making boundary preservation a persistent challenge in medical image segmentation. To address this challenge, we introduce FocusSDF, a novel loss function based on the signed distance functions (SDFs), which redirects the network to concentrate on boundary regions by adaptively assigning higher weights to pixels closer to the lesion or organ boundary, effectively making it boundary aware. To rigorously validate FocusSDF, we perform extensive evaluations against five state-of-the-art medical image segmentation models, including the foundation model MedSAM, using four distance-based loss functions across diverse datasets covering cerebral aneurysm, stroke, liver, and breast tumor segmentation tasks spanning multiple imaging modalities. The experimental results consistently demonstrate the superior performance of FocusSDF over existing distance transform based loss functions. --- ### Modeling X-ray photon pile-up with a normalizing flow Authors: Ole König, Daniela Huppenkothen, Douglas Finkbeiner, Christian Kirsch, Jörn Wilms, Justina R. Yang, James F. Steiner, Juan Rafael Martínez-Galarza Published: 2025-11-14 Categories: astro-ph.HE, astro-ph.IM, cs.LG arXiv: https://arxiv.org/abs/2511.11863v1 The dynamic range of imaging detectors flown on-board X-ray observatories often only covers a limited flux range of extrasolar X-ray sources. The analysis of bright X-ray sources is complicated by so-called pile-up, which results from high incident photon flux. This nonlinear effect distorts the measured spectrum, resulting in biases in the inferred physical parameters, and can even lead to a complete signal loss in extreme cases. Piled-up data are commonly discarded due to resulting intractability of the likelihood. As a result, a large number of archival observations remain underexplored. We present a machine learning solution to this problem, using a simulation-based inference framework that allows us to estimate posterior distributions of physical source parameters from piled-up eROSITA data. We show that a normalizing flow produces better-constrained posterior densities than traditional mitigation techniques, as more data can be leveraged. We consider model- and calibration-dependent uncertainties and the applicability of such an algorithm to real data in the eROSITA archive. --- ### Three Stage Narrative Analysis; Plot-Sentiment Breakdown, Structure Learning and Concept Detection Authors: Taimur Khan, Ramoza Ahsan, Mohib Hameed Published: 2025-11-14 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2511.11857v1 Story understanding and analysis have long been challenging areas within Natural Language Understanding. Automated narrative analysis requires deep computational semantic representations along with syntactic processing. Moreover, the large volume of narrative data demands automated semantic analysis and computational learning rather than manual analytical approaches. In this paper, we propose a framework that analyzes the sentiment arcs of movie scripts and performs extended analysis related to the context of the characters involved. The framework enables the extraction of high-level and low-level concepts conveyed through the narrative. Using dictionary-based sentiment analysis, our approach applies a custom lexicon built with the LabMTsimple storylab module. The custom lexicon is based on the Valence, Arousal, and Dominance scores from the NRC-VAD dataset. Furthermore, the framework advances the analysis by clustering similar sentiment plots using Wards hierarchical clustering technique. Experimental evaluation on a movie dataset shows that the resulting analysis is helpful to consumers and readers when selecting a narrative or story. --- ### Defending Unauthorized Model Merging via Dual-Stage Weight Protection Authors: Wei-Jia Chen, Min-Yen Tsai, Cheng-Yi Lee, Chia-Mu Yu Published: 2025-11-14 Categories: cs.CV, cs.CR arXiv: https://arxiv.org/abs/2511.11851v3 The rapid proliferation of pretrained models and open repositories has made model merging a convenient yet risky practice, allowing free-riders to combine fine-tuned models into a new multi-capability model without authorization. Such unauthorized model merging not only violates intellectual property rights but also undermines model ownership and accountability. To address this issue, we present MergeGuard, a proactive dual-stage weight protection framework that disrupts merging compatibility while maintaining task fidelity. In the first stage, we redistribute task-relevant information across layers via L2-regularized optimization, ensuring that important gradients are evenly dispersed. In the second stage, we inject structured perturbations to misalign task subspaces, breaking curvature compatibility in the loss landscape. Together, these stages reshape the model's parameter geometry such that merged models collapse into destructive interference while the protected model remains fully functional. Extensive experiments on both vision (ViT-L-14) and language (Llama2, Gemma2, Mistral) models demonstrate that MergeGuard reduces merged model accuracy by up to 90% with less than 1.5% performance loss on the protected model. --- ### Leveraging Exogenous Signals for Hydrology Time Series Forecasting Authors: Junyang He, Judy Fox, Alireza Jafari, Ying-Jung Chen, Geoffrey Fox Published: 2025-11-14 Categories: cs.LG arXiv: https://arxiv.org/abs/2511.11849v2 Recent advances in time series research facilitate the development of foundation models. While many state-of-the-art time series foundation models have been introduced, few studies examine their effectiveness in specific downstream applications in physical science. This work investigates the role of integrating domain knowledge into time series models for hydrological rainfall-runoff modeling. Using the CAMELS-US dataset, which includes rainfall and runoff data from 671 locations with six time series streams and 30 static features, we compare baseline and foundation models. Results demonstrate that models incorporating comprehensive known exogenous inputs outperform more limited approaches, including foundation models. Notably, incorporating natural annual periodic time series contribute the most significant improvements. --- ### Phase-Coded Memory and Morphological Resonance: A Next-Generation Retrieval-Augmented Generator Architecture Authors: Denis V. Saklakov Published: 2025-11-14 Categories: cs.NE arXiv: https://arxiv.org/abs/2511.11848v1 This paper introduces a cognitive Retrieval-Augmented Generator (RAG) architecture that transcends transformer context-length limitations through phase-coded memory and morphological-semantic resonance. Instead of token embeddings, the system encodes meaning as complex wave patterns with amplitude-phase structure. A three-tier design is presented: a Morphological Mapper that transforms inputs into semantic waveforms, a Field Memory Layer that stores knowledge as distributed holographic traces and retrieves it via phase interference, and a Non-Contextual Generator that produces coherent output guided by resonance rather than fixed context. This approach eliminates sequential token dependence, greatly reduces memory and computational overhead, and enables unlimited effective context through frequency-based semantic access. The paper outlines theoretical foundations, pseudocode implementation, and experimental evidence from related complex-valued neural models, emphasizing substantial energy, storage, and time savings. --- ### A Multimodal Manufacturing Safety Chatbot: Knowledge Base Design, Benchmark Development, and Evaluation of Multiple RAG Approaches Authors: Ryan Singh, Austin Hamilton, Amanda White, Michael Wise, Ibrahim Yousif, Arthur Carvalho, Zhe Shan, Reza Abrisham Baf, Mohammad Mayyas, Lora A. Cavuoto, Fadel M. Megahed Published: 2025-11-14 Categories: cs.IR, cs.AI, cs.CY arXiv: https://arxiv.org/abs/2511.11847v2 Ensuring worker safety remains a critical challenge in modern manufacturing environments. Industry 5.0 reorients the prevailing manufacturing paradigm toward more human-centric operations. Using a design science research methodology, we identify three essential requirements for next-generation safety training systems: high accuracy, low latency, and low cost. We introduce a multimodal chatbot powered by large language models that meets these design requirements. The chatbot uses retrieval-augmented generation to ground its responses in curated regulatory and technical documentation. To evaluate our solution, we developed a domain-specific benchmark of expert-validated question and answer pairs for three representative machines: a Bridgeport manual mill, a Haas TL-1 CNC lathe, and a Universal Robots UR5e collaborative robot. We tested 24 RAG configurations using a full-factorial design and assessed them with automated evaluations of correctness, latency, and cost. Our top 2 configurations were then evaluated by ten industry experts and academic researchers. Our results show that retrieval strategy and model configuration have a significant impact on performance. The top configuration, selected for chatbot deployment, achieved an accuracy of 86.66%, an average cost of $0.005 per query, and an average end-to-end latency of 10.04 seconds. This latency is practical for delivering a complete safety instruction and is measured from query submission to full instruction delivery rather than generation onset. Overall, our work provides three contributions: an open-source, domain-grounded safety training chatbot; a validated benchmark for evaluating AI-assisted safety instruction; and a systematic methodology for designing and assessing AI-enabled instructional and immersive safety training systems for Industry 5.0 environments. --- ### Autonomous Underwater Cognitive System for Adaptive Navigation: A SLAM-Integrated Cognitive Architecture Authors: K. A. I. N Jayarathne, R. M. N. M. Rathnayaka, D. P. S. S. Peiris Published: 2025-11-14 Categories: cs.RO, cs.AI, cs.AR arXiv: https://arxiv.org/abs/2511.11845v1 Deep-sea exploration poses significant challenges, including disorientation, communication loss, and navigational failures in dynamic underwater environments. This paper presents an Autonomous Underwater Cognitive System (AUCS) that integrates Simultaneous Localization and Mapping (SLAM) with a Soar-based cognitive architecture to enable adaptive navigation in complex oceanic conditions. The system fuses multi-sensor data from SONAR, LiDAR, IMU, and DVL with cognitive reasoning modules for perception, attention, planning, and learning. Unlike conventional SLAM systems, AUCS incorporates semantic understanding, adaptive sensor management, and memory-based learning to differentiate between dynamic and static objects, reducing false loop closures and enhancing long-term map consistency. The proposed architecture demonstrates a complete perception-cognition-action-learning loop, allowing autonomous underwater vehicles to sense, reason, and adapt intelligently. This work lays a foundation for next-generation cognitive submersible systems, improving safety, reliability, and autonomy in deep-sea exploration. ---