# Research Papers — 2025-12 · bundle page 91 of 175

> Full-text markdown of papers #4501–4550 of 8703 published in 2025-12, newest first. AI-readable mirrors of the full 8703-paper 2025-12 archive (50 papers per file).

Canonical page: https://aipod.dev/papers/archive/2025-12/page/91
[← Previous bundle](https://aipod.dev/papers/archive/2025-12/page/90.md)
[Next bundle →](https://aipod.dev/papers/archive/2025-12/page/92.md)

### From Tokens to Photons: Test-Time Physical Prompting for Vison-Language Models

Authors: Boyeong Im, Wooseok Lee, Yoojin Kwon, Hyung-Sin Kim
Published: 2025-12-14
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.12571v1

To extend the application of vision-language models (VLMs) from web images to sensor-mediated physical environments, we propose Multi-View Physical-prompt for Test-Time Adaptation (MVP), a forward-only framework that moves test-time adaptation (TTA) from tokens to photons by treating the camera exposure triangle--ISO, shutter speed, and aperture--as physical prompts. At inference, MVP acquires a library of physical views per scene, selects the top-k sensor settings using a source-affinity score, evaluates each retained view under lightweight digital augmentations, filters the lowest-entropy subset of augmented views, and aggregates predictions with Zero-temperature softmax (i.e., hard voting). This selection-then-vote design is simple, calibration-friendly, and requires no gradients or model modifications. On ImageNet-ES and ImageNet-ES-Diverse, MVP consistently outperforms digital-only TTA on single Auto-Exposure captures, by up to 25.6 percentage points (pp), and delivers up to 3.4 pp additional gains over pipelines that combine conventional sensor control with TTA. MVP remains effective under reduced parameter candidate sets that lower capture latency, demonstrating practicality. These results support the main claim that, beyond post-capture prompting, measurement-time control--selecting and combining real physical views--substantially improves robustness for VLMs.

---

### Optimal Mistake Bounds for Transductive Online Learning

Authors: Zachary Chase, Steve Hanneke, Shay Moran, Jonathan Shafer
Published: 2025-12-14
Categories: cs.LG, math.CO, stat.ML
arXiv: https://arxiv.org/abs/2512.12567v1

We resolve a 30-year-old open problem concerning the power of unlabeled data in online learning by tightly quantifying the gap between transductive and standard online learning. In the standard setting, the optimal mistake bound is characterized by the Littlestone dimension $d$ of the concept class $H$ (Littlestone 1987). We prove that in the transductive setting, the mistake bound is at least $Ω(\sqrt{d})$. This constitutes an exponential improvement over previous lower bounds of $Ω(\log\log d)$, $Ω(\sqrt{\log d})$, and $Ω(\log d)$, due respectively to Ben-David, Kushilevitz, and Mansour (1995, 1997) and Hanneke, Moran, and Shafer (2023). We also show that this lower bound is tight: for every $d$, there exists a class of Littlestone dimension $d$ with transductive mistake bound $O(\sqrt{d})$. Our upper bound also improves upon the best known upper bound of $(2/3)d$ from Ben-David, Kushilevitz, and Mansour (1997). These results establish a quadratic gap between transductive and standard online learning, thereby highlighting the benefit of advance access to the unlabeled instance sequence. This contrasts with the PAC setting, where transductive and standard learning exhibit similar sample complexities.

---

### StreamingAssistant: Efficient Visual Token Pruning for Accelerating Online Video Understanding

Authors: Xinqi Jin, Hanxun Yu, Bohan Yu, Kebin Liu, Jian Liu, Keda Tao, Yixuan Pei, Huan Wang, Fan Dang, Jiangchuan Liu, Weiqiang Wang
Published: 2025-12-14
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2512.12560v1

Online video understanding is essential for applications like public surveillance and AI glasses. However, applying Multimodal Large Language Models (MLLMs) to this domain is challenging due to the large number of video frames, resulting in high GPU memory usage and computational latency. To address these challenges, we propose token pruning as a means to reduce context length while retaining critical information. Specifically, we introduce a novel redundancy metric, Maximum Similarity to Spatially Adjacent Video Tokens (MSSAVT), which accounts for both token similarity and spatial position. To mitigate the bidirectional dependency between pruning and redundancy, we further design a masked pruning strategy that ensures only mutually unadjacent tokens are pruned. We also integrate an existing temporal redundancy-based pruning method to eliminate temporal redundancy of the video modality. Experimental results on multiple online and offline video understanding benchmarks demonstrate that our method significantly improves the accuracy (i.e., by 4\% at most) while incurring a negligible pruning latency (i.e., less than 1ms). Our full implementation will be made publicly available.

---

### Large Language Newsvendor: Decision Biases and Cognitive Mechanisms

Authors: Jifei Liu, Zhi Chen, Yuanguang Zhong
Published: 2025-12-14
Categories: cs.AI
arXiv: https://arxiv.org/abs/2512.12552v1

Problem definition: Although large language models (LLMs) are increasingly integrated into business decision making, their potential to replicate and even amplify human cognitive biases cautions a significant, yet not well-understood, risk. This is particularly critical in high-stakes operational contexts like supply chain management. To address this, we investigate the decision-making patterns of leading LLMs using the canonical newsvendor problem in a dynamic setting, aiming to identify the nature and origins of their cognitive biases. Methodology/results: Through dynamic, multi-round experiments with GPT-4, GPT-4o, and LLaMA-8B, we tested for five established decision biases. We found that LLMs consistently replicated the classic ``Too Low/Too High'' ordering bias and significantly amplified other tendencies like demand-chasing behavior compared to human benchmarks. Our analysis uncovered a ``paradox of intelligence'': the more sophisticated GPT-4 demonstrated the greatest irrationality through overthinking, while the efficiency-optimized GPT-4o performed near-optimally. Because these biases persist even when optimal formulas are provided, we conclude they stem from architectural constraints rather than knowledge gaps. Managerial implications: First, managers should select models based on the specific task, as our results show that efficiency-optimized models can outperform more complex ones on certain optimization problems. Second, the significant amplification of bias by LLMs highlights the urgent need for robust human-in-the-loop oversight in high-stakes decisions to prevent costly errors. Third, our findings suggest that designing structured, rule-based prompts is a practical and effective strategy for managers to constrain models' heuristic tendencies and improve the reliability of AI-assisted decisions.

---

### Iterative Sampling Methods for Sinkhorn Distributionally Robust Optimization

Authors: Jie Wang
Published: 2025-12-14
Categories: stat.ML, cs.LG
arXiv: https://arxiv.org/abs/2512.12550v1

Distributionally robust optimization (DRO) has emerged as a powerful paradigm for reliable decision-making under uncertainty. This paper focuses on DRO with ambiguity sets defined via the Sinkhorn discrepancy: an entropy-regularized Wasserstein distance, referred to as Sinkhorn DRO. Existing work primarily addresses Sinkhorn DRO from a dual perspective, leveraging its formulation as a conditional stochastic optimization problem, for which many stochastic gradient methods are applicable. However, the theoretical analyses of such methods often rely on the boundedness of the loss function, and it is indirect to obtain the worst-case distribution associated with Sinkhorn DRO. In contrast, we study Sinkhorn DRO from the primal perspective, by reformulating it as a bilevel program with several infinite-dimensional lower-level subproblems over probability space. This formulation enables us to simultaneously obtain the optimal robust decision and the worst-case distribution, which is valuable in practical settings, such as generating stress-test scenarios or designing robust learning algorithms. We propose both double-loop and single-loop sampling-based algorithms with theoretical guarantees to solve this bilevel program. Finally, we demonstrate the effectiveness of our approach through a numerical study on adversarial classification.

---

### Supervised Contrastive Frame Aggregation for Video Representation Learning

Authors: Shaif Chowdhury, Mushfika Rahman, Greg Hamerly
Published: 2025-12-14
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2512.12549v1

We propose a supervised contrastive learning framework for video representation learning that leverages temporally global context. We introduce a video to image aggregation strategy that spatially arranges multiple frames from each video into a single input image. This design enables the use of pre trained convolutional neural network backbones such as ResNet50 and avoids the computational overhead of complex video transformer models. We then design a contrastive learning objective that directly compares pairwise projections generated by the model. Positive pairs are defined as projections from videos sharing the same label while all other projections are treated as negatives. Multiple natural views of the same video are created using different temporal frame samplings from the same underlying video. Rather than relying on data augmentation these frame level variations produce diverse positive samples with global context and reduce overfitting. Experiments on the Penn Action and HMDB51 datasets demonstrate that the proposed method outperforms existing approaches in classification accuracy while requiring fewer computational resources. The proposed Supervised Contrastive Frame Aggregation method learns effective video representations in both supervised and self supervised settings and supports video based tasks such as classification and captioning. The method achieves seventy six percent classification accuracy on Penn Action compared to forty three percent achieved by ViVIT and forty eight percent accuracy on HMDB51 compared to thirty seven percent achieved by ViVIT.

---

### World Models Unlock Optimal Foraging Strategies in Reinforcement Learning Agents

Authors: Yesid Fonseca, Manuel S. Ríos, Nicanor Quijano, Luis F. Giraldo
Published: 2025-12-14
Categories: cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2512.12548v2

Patch foraging involves the deliberate and planned process of determining the optimal time to depart from a resource-rich region and investigate potentially more beneficial alternatives. The Marginal Value Theorem (MVT) is frequently used to characterize this process, offering an optimality model for such foraging behaviors. Although this model has been widely used to make predictions in behavioral ecology, discovering the computational mechanisms that facilitate the emergence of optimal patch-foraging decisions in biological foragers remains under investigation. Here, we show that artificial foragers equipped with learned world models naturally converge to MVT-aligned strategies. Using a model-based reinforcement learning agent that acquires a parsimonious predictive representation of its environment, we demonstrate that anticipatory capabilities, rather than reward maximization alone, drive efficient patch-leaving behavior. Compared with standard model-free RL agents, these model-based agents exhibit decision patterns similar to many of their biological counterparts, suggesting that predictive world models can serve as a foundation for more explainable and biologically grounded decision-making in AI systems. Overall, our findings highlight the value of ecological optimality principles for advancing interpretable and adaptive AI.

---

### World Models Unlock Optimal Foraging Strategies in Reinforcement Learning Agents

Authors: Yesid Fonseca, Manuel S. Ríos, Nicanor Quijano, Luis F. Giraldo
Published: 2025-12-14
Categories: cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2512.12548v1

Patch foraging involves the deliberate and planned process of determining the optimal time to depart from a resource-rich region and investigate potentially more beneficial alternatives. The Marginal Value Theorem (MVT) is frequently used to characterize this process, offering an optimality model for such foraging behaviors. Although this model has been widely used to make predictions in behavioral ecology, discovering the computational mechanisms that facilitate the emergence of optimal patch-foraging decisions in biological foragers remains under investigation. Here, we show that artificial foragers equipped with learned world models naturally converge to MVT-aligned strategies. Using a model-based reinforcement learning agent that acquires a parsimonious predictive representation of its environment, we demonstrate that anticipatory capabilities, rather than reward maximization alone, drive efficient patch-leaving behavior. Compared with standard model-free RL agents, these model-based agents exhibit decision patterns similar to many of their biological counterparts, suggesting that predictive world models can serve as a foundation for more explainable and biologically grounded decision-making in AI systems. Overall, our findings highlight the value of ecological optimality principles for advancing interpretable and adaptive AI.

---

### Skillful Subseasonal-to-Seasonal Forecasting of Extreme Events with a Multi-Sphere Coupled Probabilistic Model

Authors: Bin Mu, Yuxuan Chen, Shijin Yuan, Bo Qin, Hao Guo
Published: 2025-12-14
Categories: cs.LG, cs.AI, physics.ao-ph
arXiv: https://arxiv.org/abs/2512.12545v1

Accurate subseasonal-to-seasonal (S2S) prediction of extreme events is critical for resource planning and disaster mitigation under accelerating climate change. However, such predictions remain challenging due to complex multi-sphere interactions and intrinsic atmospheric uncertainty. Here we present TianXing-S2S, a multi-sphere coupled probabilistic model for global S2S daily ensemble forecast. TianXing-S2S first encodes diverse multi-sphere predictors into a compact latent space, then employs a diffusion model to generate daily ensemble forecasts. A novel coupling module based on optimal transport (OT) is incorporated in the denoiser to optimize the interactions between atmospheric and multi-sphere boundary conditions. Across key atmospheric variables, TianXing-S2S outperforms both the European Centre for Medium-Range Weather Forecasts (ECMWF) S2S system and FuXi-S2S in 45-day daily-mean ensemble forecasts at 1.5 resolution. Our model achieves skillful subseasonal prediction of extreme events including heat waves and anomalous precipitation, identifying soil moisture as a critical precursor signal. Furthermore, we demonstrate that TianXing-S2S can generate stable rollout forecasts up to 180 days, establishing a robust framework for S2S research in a warming world.

---

### HyperEdit: Unlocking Instruction-based Text Editing in LLMs via Hypernetworks

Authors: Yiming Zeng, Jinghan Cao, Zexin Li, Wanhao Yu, Zhankai Ye, Dawei Xiang, Ting Hua, Xin Liu, Shangqian Gao, Tingting Yu
Published: 2025-12-14
Categories: cs.CL, cs.LG
arXiv: https://arxiv.org/abs/2512.12544v1

Instruction-based text editing is increasingly critical for real-world applications such as code editors (e.g., Cursor), but Large Language Models (LLMs) continue to struggle with this task. Unlike free-form generation, editing requires faithfully implementing user instructions while preserving unchanged content, as even minor unintended modifications can break functionality. Existing approaches treat editing as generic text generation, leading to two key failures: they struggle to faithfully align edits with diverse user intents, and they often over-edit unchanged regions. We propose HyperEdit to address both issues. First, we introduce hypernetwork-based dynamic adaptation that generates request-specific parameters, enabling the model to tailor its editing strategy to each instruction. Second, we develop difference-aware regularization that focuses supervision on modified spans, preventing over-editing while ensuring precise, minimal changes. HyperEdit achieves a 9%--30% relative improvement in BLEU on modified regions over state-of-the-art baselines, despite utilizing only 3B parameters.

---

### Effective Fine-Tuning with Eigenvector Centrality Based Pruning

Authors: Shaif Chowdhury, Soham Biren Katlariwala, Devleena Kashyap
Published: 2025-12-14
Categories: cs.LG
arXiv: https://arxiv.org/abs/2512.12543v1

In social media networks a small number of highly influential users can drive large scale changes in discourse across multiple communities. Small shifts in the behavior of these users are often sufficient to propagate widely throughout the network. A similar phenomenon occurs during neural network fine tuning. Conventional fine tuning of convolutional neural networks typically adds a new linear classification layer on top of a large pre trained model. Instead we argue that improved adaptation can be achieved by first pruning the network to retain only the most important neurons and then performing fine tuning. We propose a graph theory based method for pruning neural networks that is designed to improve fine tuning performance. In this method each neuron is represented as a node and edges encode similarity between neurons. Neurons are pruned based on importance scores computed using eigenvector centrality. The resulting pruned network is then fine tuned using only the most central neurons. We evaluate the proposed method on VGGNet EfficientNet and ResNet models using the TF Flowers Caltech one zero one and Oxford Flowers one zero two datasets. The proposed approach achieves higher classification accuracy while significantly reducing model complexity. On the Oxford Flowers one zero two dataset the method achieves forty eight percent classification accuracy compared to thirty percent accuracy obtained by the baseline VGGNet model.

---

### Anatomy Guided Coronary Artery Segmentation from CCTA Using Spatial Frequency Joint Modeling

Authors: Huan Huang, Michele Esposito, Chen Zhao
Published: 2025-12-14
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.12539v1

Accurate coronary artery segmentation from coronary computed tomography angiography is essential for quantitative coronary analysis and clinical decision support. Nevertheless, reliable segmentation remains challenging because of small vessel calibers, complex branching, blurred boundaries, and myocardial interference. We propose a coronary artery segmentation framework that integrates myocardial anatomical priors, structure aware feature encoding, and three dimensional wavelet inverse wavelet transformations. Myocardial priors and residual attention based feature enhancement are incorporated during encoding to strengthen coronary structure representation. Wavelet inverse wavelet based downsampling and upsampling enable joint spatial frequency modeling and preserve multi scale structural consistency, while a multi scale feature fusion module integrates semantic and geometric information in the decoding stage. The model is trained and evaluated on the public ImageCAS dataset using a 3D overlapping patch based strategy with a 7:1:2 split for training, validation, and testing. Experimental results demonstrate that the proposed method achieves a Dice coefficient of 0.8082, Sensitivity of 0.7946, Precision of 0.8471, and an HD95 of 9.77 mm, outperforming several mainstream segmentation models. Ablation studies further confirm the complementary contributions of individual components. The proposed method enables more stable and consistent coronary artery segmentation under complex geometric conditions, providing reliable segmentation results for subsequent coronary structure analysis tasks.

---

### NagaNLP: Bootstrapping NLP for Low-Resource Nagamese Creole with Human-in-the-Loop Synthetic Data

Authors: Agniva Maiti, Manya Pandey, Murari Mandal
Published: 2025-12-14
Categories: cs.CL
arXiv: https://arxiv.org/abs/2512.12537v1

The vast majority of the world's languages, particularly creoles like Nagamese, remain severely under-resourced in Natural Language Processing (NLP), creating a significant barrier to their representation in digital technology. This paper introduces NagaNLP, a comprehensive open-source toolkit for Nagamese, bootstrapped through a novel methodology that relies on LLM-driven but human-validated synthetic data generation. We detail a multi-stage pipeline where an expert-guided LLM (Gemini) generates a candidate corpus, which is then refined and annotated by native speakers. This synthetic-hybrid approach yielded a 10K pair conversational dataset and a high-quality annotated corpus for foundational tasks. To assess the effectiveness of our methodology, we trained both discriminative and generative models. Our fine-tuned XLM-RoBERTa-base model establishes a new benchmark for Nagamese, achieving a 93.81\% accuracy (0.90 F1-Macro) on Part-of-Speech tagging and a 0.75 F1-Macro on Named Entity Recognition, massively outperforming strong zero-shot baselines. Furthermore, we fine-tuned a Llama-3.2-3B Instruct model, named NagaLLaMA, which demonstrates superior performance on conversational tasks, achieving a Perplexity of 3.85, an order of magnitude improvement over its few-shot counterpart (96.76). We release the NagaNLP toolkit, including all datasets, models, and code, providing a foundational resource for a previously underserved language and a reproducible framework for reducing data scarcity in other low-resource contexts.

---

### Diverse LLMs vs. Vulnerabilities: Who Detects and Fixes Them Better?

Authors: Arastoo Zibaeirad, Marco Vieira
Published: 2025-12-14
Categories: cs.SE, cs.AI
arXiv: https://arxiv.org/abs/2512.12536v1

Large Language Models (LLMs) are increasingly being studied for Software Vulnerability Detection (SVD) and Repair (SVR). Individual LLMs have demonstrated code understanding abilities, but they frequently struggle when identifying complex vulnerabilities and generating fixes. This study presents DVDR-LLM, an ensemble framework that combines outputs from diverse LLMs to determine whether aggregating multiple models reduces error rates. Our evaluation reveals that DVDR-LLM achieves 10-12% higher detection accuracy compared to the average performance of individual models, with benefits increasing as code complexity grows. For multi-file vulnerabilities, the ensemble approach demonstrates significant improvements in recall (+18%) and F1 score (+11.8%) over individual models. However, the approach raises measurable trade-offs: reducing false positives in verification tasks while simultaneously increasing false negatives in detection tasks, requiring careful decision on the required level of agreement among the LLMs (threshold) for increased performance across different security contexts. Artifact: https://github.com/Erroristotle/DVDR_LLM

---

### Animus3D: Text-driven 3D Animation via Motion Score Distillation

Authors: Qi Sun, Can Wang, Jiaxiang Shang, Wensen Feng, Jing Liao
Published: 2025-12-14
Categories: cs.CV, cs.GR, cs.LG
arXiv: https://arxiv.org/abs/2512.12534v1

We present Animus3D, a text-driven 3D animation framework that generates motion field given a static 3D asset and text prompt. Previous methods mostly leverage the vanilla Score Distillation Sampling (SDS) objective to distill motion from pretrained text-to-video diffusion, leading to animations with minimal movement or noticeable jitter. To address this, our approach introduces a novel SDS alternative, Motion Score Distillation (MSD). Specifically, we introduce a LoRA-enhanced video diffusion model that defines a static source distribution rather than pure noise as in SDS, while another inversion-based noise estimation technique ensures appearance preservation when guiding motion. To further improve motion fidelity, we incorporate explicit temporal and spatial regularization terms that mitigate geometric distortions across time and space. Additionally, we propose a motion refinement module to upscale the temporal resolution and enhance fine-grained details, overcoming the fixed-resolution constraints of the underlying video model. Extensive experiments demonstrate that Animus3D successfully animates static 3D assets from diverse text prompts, generating significantly more substantial and detailed motion than state-of-the-art baselines while maintaining high visual integrity. Code will be released at https://qiisun.github.io/animus3d_page.

---

### Empirical Mode Decomposition and Graph Transformation of the MSCI World Index: A Multiscale Topological Analysis for Graph Neural Network Modeling

Authors: Agustín M. de los Riscos, Julio E. Sandubete, Diego Carmona-Fernández, León Beleña
Published: 2025-12-14
Categories: cs.LG, q-fin.CP
arXiv: https://arxiv.org/abs/2512.12526v1

This study applies Empirical Mode Decomposition (EMD) to the MSCI World index and converts the resulting intrinsic mode functions (IMFs) into graph representations to enable modeling with graph neural networks (GNNs). Using CEEMDAN, we extract nine IMFs spanning high-frequency fluctuations to long-term trends. Each IMF is transformed into a graph using four time-series-to-graph methods: natural visibility, horizontal visibility, recurrence, and transition graphs. Topological analysis shows clear scale-dependent structure: high-frequency IMFs yield dense, highly connected small-world graphs, whereas low-frequency IMFs produce sparser networks with longer characteristic path lengths. Visibility-based methods are more sensitive to amplitude variability and typically generate higher clustering, while recurrence graphs better preserve temporal dependencies. These results provide guidance for designing GNN architectures tailored to the structural properties of decomposed components, supporting more effective predictive modeling of financial time series.

---

### Noise-robust Contrastive Learning for Critical Transition Detection in Dynamical Systems

Authors: Wenqi Fang, Ye Li
Published: 2025-12-14
Categories: cs.LG, cs.AI, physics.comp-ph
arXiv: https://arxiv.org/abs/2512.12523v1

Detecting critical transitions in complex, noisy time-series data is a fundamental challenge across science and engineering. Such transitions may be anticipated by the emergence of a low-dimensional order parameter, whose signature is often masked by high-amplitude stochastic variability. Standard contrastive learning approaches based on deep neural networks, while promising for detecting critical transitions, are often overparameterized and sensitive to irrelevant noise, leading to inaccurate identification of critical points. To address these limitations, we propose a neural network architecture, constructed using singular value decomposition technique, together with a strictly semi-orthogonality-constrained training algorithm, to enhance the performance of traditional contrastive learning. Extensive experiments demonstrate that the proposed method matches the performance of traditional contrastive learning techniques in identifying critical transitions, yet is considerably more lightweight and markedly more resistant to noise.

---

### Can You Keep a Secret? Exploring AI for Care Coordination in Cognitive Decline

Authors: Alicia, Lee, Mai Lee Chang, Sreehana Mandava, Destiny Deshields, Hugo Simão, Aaron Steinfeld, Jodi Forlizzi, John Zimmerman
Published: 2025-12-14
Categories: cs.HC, cs.AI
arXiv: https://arxiv.org/abs/2512.12510v1

The increasing number of older adults who experience cognitive decline places a burden on informal caregivers, whose support with tasks of daily living determines whether older adults can remain in their homes. To explore how agents might help lower-SES older adults to age-in-place, we interviewed ten pairs of older adults experiencing cognitive decline and their informal caregivers. We explored how they coordinate care, manage burdens, and sustain autonomy and privacy. Older adults exercised control by delegating tasks to specific caregivers, keeping information about all the care they received from their adult children. Many abandoned some tasks of daily living, lowering their quality of life to ease caregiver burden. One effective strategy, piggybacking, uses spontaneous overlaps in errands to get more work done with less caregiver effort. This raises the questions: (i) Can agents help with piggyback coordination? (ii) Would it keep older adults in their homes longer, while not increasing caregiver burden?

---

### Generative Spatiotemporal Data Augmentation

Authors: Jinfan Zhou, Lixin Luo, Sungmin Eum, Heesung Kwon, Jeong Joon Park
Published: 2025-12-14
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2512.12508v1

We explore spatiotemporal data augmentation using video foundation models to diversify both camera viewpoints and scene dynamics. Unlike existing approaches based on simple geometric transforms or appearance perturbations, our method leverages off-the-shelf video diffusion models to generate realistic 3D spatial and temporal variations from a given image dataset. Incorporating these synthesized video clips as supplemental training data yields consistent performance gains in low-data settings, such as UAV-captured imagery where annotations are scarce. Beyond empirical improvements, we provide practical guidelines for (i) choosing an appropriate spatiotemporal generative setup, (ii) transferring annotations to synthetic frames, and (iii) addressing disocclusion - regions newly revealed and unlabeled in generated views. Experiments on COCO subsets and UAV-captured datasets show that, when applied judiciously, spatiotemporal augmentation broadens the data distribution along axes underrepresented by traditional and prior generative methods, offering an effective lever for improving model performance in data-scarce regimes.

---

### Explainable Artificial Intelligence for Economic Time Series: A Comprehensive Review and a Systematic Taxonomy of Methods and Concepts

Authors: Agustín García-García, Pablo Hidalgo, Julio E. Sandubete
Published: 2025-12-14
Categories: econ.GN, cs.AI
arXiv: https://arxiv.org/abs/2512.12506v1

Explainable Artificial Intelligence (XAI) is increasingly required in computational economics, where machine-learning forecasters can outperform classical econometric models but remain difficult to audit and use for policy. This survey reviews and organizes the growing literature on XAI for economic time series, where autocorrelation, non-stationarity, seasonality, mixed frequencies, and regime shifts can make standard explanation techniques unreliable or economically implausible. We propose a taxonomy that classifies methods by (i) explanation mechanism: propagation-based approaches (e.g., Integrated Gradients, Layer-wise Relevance Propagation), perturbation and game-theoretic attribution (e.g., permutation importance, LIME, SHAP), and function-based global tools (e.g., Accumulated Local Effects); (ii) time-series compatibility, including preservation of temporal dependence, stability over time, and respect for data-generating constraints. We synthesize time-series-specific adaptations such as vector- and window-based formulations (e.g., Vector SHAP, WindowSHAP) that reduce lag fragmentation and computational cost while improving interpretability. We also connect explainability to causal inference and policy analysis through interventional attributions (Causal Shapley values) and constrained counterfactual reasoning. Finally, we discuss intrinsically interpretable architectures (notably attention-based transformers) and provide guidance for decision-grade applications such as nowcasting, stress testing, and regime monitoring, emphasizing attribution uncertainty and explanation dynamics as indicators of structural change.

---

### KidsArtBench: Multi-Dimensional Children's Art Evaluation with Attribute-Aware MLLMs

Authors: Mingrui Ye, Chanjin Zheng, Zengyi Yu, Chenyu Xiang, Zhixue Zhao, Zheng Yuan, Helen Yannakoudakis
Published: 2025-12-14
Categories: cs.AI
arXiv: https://arxiv.org/abs/2512.12503v1

Multimodal Large Language Models (MLLMs) show remarkable progress across many visual-language tasks; however, their capacity to evaluate artistic expression remains limited. Aesthetic concepts are inherently abstract and open-ended, and multimodal artwork annotations are scarce. We introduce KidsArtBench, a new benchmark of over 1k children's artworks (ages 5-15) annotated by 12 expert educators across 9 rubric-aligned dimensions, together with expert comments for feedback. Unlike prior aesthetic datasets that provide single scalar scores on adult imagery, KidsArtBench targets children's artwork and pairs multi-dimensional annotations with comment supervision to enable both ordinal assessment and formative feedback. Building on this resource, we propose an attribute-specific multi-LoRA approach, where each attribute corresponds to a distinct evaluation dimension (e.g., Realism, Imagination) in the scoring rubric, with Regression-Aware Fine-Tuning (RAFT) to align predictions with ordinal scales. On Qwen2.5-VL-7B, our method increases correlation from 0.468 to 0.653, with the largest gains on perceptual dimensions and narrowed gaps on higher-order attributes. These results show that educator-aligned supervision and attribute-aware training yield pedagogically meaningful evaluations and establish a rigorous testbed for sustained progress in educational AI. We release data and code with ethics documentation.

---

### SafeGen: Embedding Ethical Safeguards in Text-to-Image Generation

Authors: Dang Phuong Nam, Nguyen Kieu, Pham Thanh Hieu
Published: 2025-12-14
Categories: cs.AI
arXiv: https://arxiv.org/abs/2512.12501v1

Generative Artificial Intelligence (AI) has created unprecedented opportunities for creative expression, education, and research. Text-to-image systems such as DALL.E, Stable Diffusion, and Midjourney can now convert ideas into visuals within seconds, but they also present a dual-use dilemma, raising critical ethical concerns: amplifying societal biases, producing high-fidelity disinformation, and violating intellectual property. This paper introduces SafeGen, a framework that embeds ethical safeguards directly into the text-to-image generation pipeline, grounding its design in established principles for Trustworthy AI. SafeGen integrates two complementary components: BGE-M3, a fine-tuned text classifier that filters harmful or misleading prompts, and Hyper-SD, an optimized diffusion model that produces high fidelity, semantically aligned images. Built on a curated multilingual (English- Vietnamese) dataset and a fairness-aware training process, SafeGen demonstrates that creative freedom and ethical responsibility can be reconciled within a single workflow. Quantitative evaluations confirm its effectiveness, with Hyper-SD achieving IS = 3.52, FID = 22.08, and SSIM = 0.79, while BGE-M3 reaches an F1-Score of 0.81. An ablation study further validates the importance of domain-specific fine-tuning for both modules. Case studies illustrate SafeGen's practical impact in blocking unsafe prompts, generating inclusive teaching materials, and reinforcing academic integrity.

---

### Explainable AI as a Double-Edged Sword in Dermatology: The Impact on Clinicians versus The Public

Authors: Xuhai Xu, Haoyu Hu, Haoran Zhang, Will Ke Wang, Reina Wang, Luis R. Soenksen, Omar Badri, Sheharbano Jafry, Elise Burger, Lotanna Nwandu, Apoorva Mehta, Erik P. Duhaime et al.
Published: 2025-12-14
Categories: cs.HC, cs.AI
arXiv: https://arxiv.org/abs/2512.12500v1

Artificial intelligence (AI) is increasingly permeating healthcare, from physician assistants to consumer applications. Since AI algorithm's opacity challenges human interaction, explainable AI (XAI) addresses this by providing AI decision-making insight, but evidence suggests XAI can paradoxically induce over-reliance or bias. We present results from two large-scale experiments (623 lay people; 153 primary care physicians, PCPs) combining a fairness-based diagnosis AI model and different XAI explanations to examine how XAI assistance, particularly multimodal large language models (LLMs), influences diagnostic performance. AI assistance balanced across skin tones improved accuracy and reduced diagnostic disparities. However, LLM explanations yielded divergent effects: lay users showed higher automation bias - accuracy boosted when AI was correct, reduced when AI erred - while experienced PCPs remained resilient, benefiting irrespective of AI accuracy. Presenting AI suggestions first also led to worse outcomes when the AI was incorrect for both groups. These findings highlight XAI's varying impact based on expertise and timing, underscoring LLMs as a "double-edged sword" in medical AI and informing future human-AI collaborative system design.

---

### Screen, Cache, and Match: A Training-Free Causality-Consistent Reference Frame Framework for Human Animation

Authors: Jianan Wang, Nailei Hei, Li He, Huanzhen Wang, Aoxing Li, Yingkai Zhao, Yuxuan Lin, Haofen Wang, Chunyang Wang, Yan Wang, Wenqiang Zhang
Published: 2025-12-13
Categories: cs.GR, cs.AI
arXiv: https://arxiv.org/abs/2601.22160v2

Human animation aims to generate temporally coherent and visually consistent videos over long sequences, yet modeling long-range dependencies while preserving frame quality remains challenging. Inspired by the human ability to leverage past observations for interpreting ongoing actions, we propose FrameCache, a training-free, causality-consistent reference frame framework. FrameCache explicitly converts historical generation results into causal guidance through two complementary mechanisms. First, at the reference level, a novel Screen-Cache-Match (SCM) strategy constructs a dynamic, high-quality reference memory, ensuring motion-consistent appearance guidance to reduce identity drift. Second, at the generative level, a Trajectory-Aware Autoregressive Generation (TAAG) mechanism aligns denoising trajectories across adjacent video chunks. This is achieved through an overlap-aware latent propagation and a dual-domain fusion strategy that seamlessly blends low-frequency structural layouts with high-frequency textural details. Extensive experiments on standard benchmarks demonstrate that FrameCache consistently improves temporal coherence and visual stability while integrating seamlessly with diverse diffusion baselines. Code will be made publicly available.

---

### On Harnessing Idle Compute at the Edge for Foundation Model Training

Authors: Leyang Xue, Meghana Madhyastha, Myungjin Lee, Amos Storkey, Randal Burns, Mahesh K. Marina
Published: 2025-12-13
Categories: cs.DC, cs.LG
arXiv: https://arxiv.org/abs/2512.22142v2

The foundation-model ecosystem remains highly centralized because training requires immense compute resources and is therefore largely limited to large cloud operators. Edge-assisted foundation model training that harnesses spare compute on edge devices offers a more democratized alternative. However, existing edge-training approaches fall short: they struggle to match cloud-training performance, scale to larger models, fit within device memory limits, or keep communication overhead manageable. They also do not handle device heterogeneity and churn satisfactorily. We introduce Cleave, built on a structural insight: each GEMM has an asymmetric I/O pattern -- its input matrices, sent over downlink, are much larger than the partial output blocks returned over uplink -- matching edge networks where downlink bandwidth exceeds uplink by 2--10x. Exploiting this alignment with a parameter-server-centric architecture, Cleave makes per-device communication \emph{decrease} as more devices join, rather than stay constant as in conventional TP. Decomposing training into independent sub-GEMM tasks yields one scheduling abstraction that unifies memory constraints, communication overhead, and fault tolerance under device churn. Our evaluation shows that Cleave achieves cloud-comparable GPU training performance and outperforms state-of-the-art edge-training methods by 4--10x in per-batch runtime at the same device counts. Beyond this shared operating range, Cleave scales to thousands of heterogeneous devices -- a regime where prior edge-training systems cannot operate -- and achieves at least 100x faster recovery from device failures.

---

### Bidirectional human-AI collaboration in brain tumour assessments improves both expert human and AI agent performance

Authors: James K Ruffle, Samia Mohinta, Guilherme Pombo, Asthik Biswas, Alan Campbell, Indran Davagnanam, David Doig, Ahmed Hammam, Harpreet Hyare, Farrah Jabeen, Emma Lim, Dermot Mallon et al.
Published: 2025-12-13
Categories: cs.HC, cs.AI, cs.LG, cs.MA
arXiv: https://arxiv.org/abs/2512.19707v2

The benefits of artificial intelligence (AI) human partnerships-evaluating how AI agents enhance expert human performance-are increasingly studied. Though rarely evaluated in healthcare, an inverse approach is possible: AI benefiting from the support of an expert human agent. Here, we investigate both human-AI clinical partnership paradigms in the magnetic resonance imaging-guided characterisation of patients with brain tumours. We reveal that human-AI partnerships improve accuracy and metacognitive ability not only for radiologists supported by AI, but also for AI agents supported by radiologists. Moreover, the greatest patient benefit was evident with an AI agent supported by a human one. Synergistic improvements in agent accuracy, metacognitive performance, and inter-rater agreement suggest that AI can create more capable, confident, and consistent clinical agents, whether human or model-based. Our work suggests that the maximal value of AI in healthcare could emerge not from replacing human intelligence, but from AI agents that routinely leverage and amplify it.

---

### NystagmusNet: Explainable Deep Learning for Photosensitivity Risk Prediction

Authors: Karthik Prabhakar
Published: 2025-12-13
Categories: cs.CV, cs.AI
arXiv: https://arxiv.org/abs/2512.17943v1

Nystagmus patients with photosensitivity face significant daily challenges due to involuntary eye movements exacerbated by environmental brightness conditions. Current assistive solutions are limited to symptomatic treatments without predictive personalization. This paper proposes NystagmusNet, an AI-driven system that predicts high-risk visual environments and recommends real-time visual adaptations. Using a dual-branch convolutional neural network trained on synthetic and augmented datasets, the system estimates a photosensitivity risk score based on environmental brightness and eye movement variance. The model achieves 75% validation accuracy on synthetic data. Explainability techniques including SHAP and GradCAM are integrated to highlight environmental risk zones, improving clinical trust and model interpretability. The system includes a rule-based recommendation engine for adaptive filter suggestions. Future directions include deployment via smart glasses and reinforcement learning for personalized recommendations.

---

### Accelerated Digital Twin Learning for Edge AI: A Comparison of FPGA and Mobile GPU

Authors: Bin Xu, Ayan Banerjee, Midhat Urooj, Sandeep K. S. Gupta
Published: 2025-12-13
Categories: cs.DC, cs.AI
arXiv: https://arxiv.org/abs/2512.17941v1

Digital twins (DTs) can enable precision healthcare by continually learning a mathematical representation of patient-specific dynamics. However, mission critical healthcare applications require fast, resource-efficient DT learning, which is often infeasible with existing model recovery (MR) techniques due to their reliance on iterative solvers and high compute/memory demands. In this paper, we present a general DT learning framework that is amenable to acceleration on reconfigurable hardware such as FPGAs, enabling substantial speedup and energy efficiency. We compare our FPGA-based implementation with a multi-processing implementation in mobile GPU, which is a popular choice for AI in edge devices. Further, we compare both edge AI implementations with cloud GPU baseline. Specifically, our FPGA implementation achieves an 8.8x improvement in \text{performance-per-watt} for the MR task, a 28.5x reduction in DRAM footprint, and a 1.67x runtime speedup compared to cloud GPU baselines. On the other hand, mobile GPU achieves 2x better performance per watts but has 2x increase in runtime and 10x more DRAM footprint than FPGA. We show the usage of this technique in DT guided synthetic data generation for Type 1 Diabetes and proactive coronary artery disease detection.

---

### Emergence: Overcoming Privileged Information Bias in Asymmetric Embodied Agents via Active Querying

Authors: Shaun Baek, Sam Liu, Joseph Ukpong
Published: 2025-12-13
Categories: cs.AI, cs.MA, cs.RO
arXiv: https://arxiv.org/abs/2512.15776v1

Large Language Models (LLMs) act as powerful reasoning engines but struggle with "symbol grounding" in embodied environments, particularly when information is asymmetrically distributed. We investigate the Privileged Information Bias (or "Curse of Knowledge"), where a knowledgeable "Leader" agent fails to guide a sensor-limited "Follower" due to a lack of Theory of Mind. To quantify this phenomenon, we propose a novel Asymmetric Assistive Reasoning framework within AI2-THOR. Our experiments reveal a significant "Success Gap": while the Leader successfully perceives the target in 35.0% of episodes, the collaborative team succeeds only 17.0% of the time, implying that nearly 50% of feasible plans fail solely due to communicative grounding errors. We demonstrate that a "Pull-based" protocol (active querying) is significantly more robust than standard "Push-based" instruction, with successful episodes featuring 2x the frequency of clarification requests. This research isolates the mechanism of active uncertainty reduction as a prerequisite for safe human-AI and robot-robot collaboration.

---

### Enhanced Web User Interface Design Via Cross-Device Responsiveness Assessment Using An Improved HCI-INTEGRATED DL Schemes

Authors: Shrinivass Arunachalam Balasubramanian
Published: 2025-12-13
Categories: cs.HC, cs.AI, cs.LG, cs.MM, cs.SE
arXiv: https://arxiv.org/abs/2512.15775v1

User Interface (UI) optimization is essential in the digital era to enhance user satisfaction in web environments. Nevertheless, the existing UI optimization models had overlooked the Cross-Responsiveness (CR) assessment, affecting the user interaction efficiency. Consequently, this article proposes a dynamic web UI optimization through CR assessment using Finite Exponential Continuous State Machine (FECSM) and Quokka Nonlinear Difference Swarm Optimization Algorithm (QNDSOA). Initially, the design and user interaction related information is collected as well as pre-processed for min-max normalization. Next, the Human-Computer Interaction (HCI)-based features are extracted, followed by user behaviour pattern grouping. Meanwhile, the CR assessment is done using FECSM. Then, the proposed Bidirectional Gated Luong and Mish Recurrent Unit (BiGLMRU) is used to classify the User eXperience (UX) change type, which is labelled based on the User Interface Change Prediction Index (UICPI). Lastly, a novel QNDSOA is utilized to optimize the UI design with an average fitness of 98.5632%. Feedback monitoring is done after optimal deployment.

---

### Two-Step Data Augmentation for Masked Face Detection and Recognition: Turning Fake Masks to Real

Authors: Yan Yang, George Bebis, Mircea Nicolescu
Published: 2025-12-13
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2512.15774v5

The absence of large-scale masked face datasets challenges masked face detection and recognition. We propose a two-step generative data augmentation framework combining rule-based mask warping with unpaired image-to-image translation via GANs, producing masked face samples that go beyond rule-based overlays. Trained on about 19,100 images in the target domain (3.8% of IAMGAN's scale), or, including out-of-domain transfer pretraining, 59,600 and 11.8%, the proposed approach yields consistent improvements over rule-based warping alone and achieves results complementary to IAMGAN's, showing that both steps contribute. Evaluation is conducted directly on the generated samples and is qualitative; quantitative metrics like FID and KID were not applied as any real reference distribution would unfairly favor the model with closer training data. We introduce a non-mask preservation loss to reduce non-mask distortions and stabilize training, and stochastic noise injection to enhance sample diversity. Note: The paper originated as a coursework project completed under resource constraints. Following scholarship termination, the author took on part-time employment to maintain research continuity, which led to a mid-semester domain pivot from medical imaging to masked face tasks due to company data restrictions. The work was completed alongside concurrent coursework with delayed compute access and without AI assistance. It was submitted at the semester end to meet a publication requirement, accepted without revision requests, and selected among the best papers for an extended submission to Springer Nature Computer Science, which was not pursued due to continued funding absence. Downstream evaluation on recognition or detection performance was not completed by the submission deadline. The note is added in response to subsequent comparisons and criticisms that did not account for these conditions.

---

### TS-DP: Reinforcement Speculative Decoding For Temporal Adaptive Diffusion Policy Acceleration

Authors: Ye Li, Jiahe Feng, Yuan Meng, Kangye Ji, Chen Tang, Xinwan Wen, Shutao Xia, Zhi Wang, Wenwu Zhu
Published: 2025-12-13
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2512.15773v1

Diffusion Policy (DP) excels in embodied control but suffers from high inference latency and computational cost due to multiple iterative denoising steps. The temporal complexity of embodied tasks demands a dynamic and adaptable computation mode. Static and lossy acceleration methods, such as quantization, fail to handle such dynamic embodied tasks, while speculative decoding offers a lossless and adaptive yet underexplored alternative for DP. However, it is non-trivial to address the following challenges: how to match the base model's denoising quality at lower cost under time-varying task difficulty in embodied settings, and how to dynamically and interactively adjust computation based on task difficulty in such environments. In this paper, we propose Temporal-aware Reinforcement-based Speculative Diffusion Policy (TS-DP), the first framework that enables speculative decoding for DP with temporal adaptivity. First, to handle dynamic environments where task difficulty varies over time, we distill a Transformer-based drafter to imitate the base model and replace its costly denoising calls. Second, an RL-based scheduler further adapts to time-varying task difficulty by adjusting speculative parameters to maintain accuracy while improving efficiency. Extensive experiments across diverse embodied environments demonstrate that TS-DP achieves up to 4.17 times faster inference with over 94% accepted drafts, reaching an inference frequency of 25 Hz and enabling real-time diffusion-based control without performance degradation.

---

### Solving PDEs With Deep Neural Nets under General Boundary Conditions

Authors: Chenggong Zhang
Published: 2025-12-13
Categories: cs.LG, cs.AI, math.NA, stat.ML
arXiv: https://arxiv.org/abs/2512.15771v3

Partial Differential Equations (PDEs) are central to modeling complex systems across physical, biological, and engineering domains, yet traditional numerical methods often struggle with high-dimensional or complex problems. Physics-Informed Neural Networks (PINNs) have emerged as an efficient alternative by embedding physics-based constraints into deep learning frameworks, but they face challenges in achieving high accuracy and handling complex boundary conditions. In this work, we extend the Time-Evolving Natural Gradient (TENG) framework to address Dirichlet boundary conditions, integrating natural gradient optimization with numerical time-stepping schemes, including Euler and Heun methods, to ensure both stability and accuracy. By incorporating boundary condition penalty terms into the loss function, the proposed approach enables precise enforcement of Dirichlet constraints. Experiments on the heat equation demonstrate the superior accuracy of the Heun method due to its second-order corrections and the computational efficiency of the Euler method for simpler scenarios. This work establishes a foundation for extending the framework to Neumann and mixed boundary conditions, as well as broader classes of PDEs, advancing the applicability of neural network-based solvers for real-world problems.

---

### Multiscale Cross-Modal Mapping of Molecular, Pathologic, and Radiologic Phenotypes in Lipid-Deficient Clear Cell Renal CellCarcinoma

Authors: Ying Cui, Dongzhe Zheng, Ke Yu, Xiyin Zheng, Xiaorui Wang, Xinxiang Li, Yan Gu, Lin Fu, Xinyi Chen, Wenjie Mei, Xin-Gui Peng
Published: 2025-12-13
Categories: q-bio.QM, cs.AI
arXiv: https://arxiv.org/abs/2512.14750v1

Clear cell renal cell carcinoma (ccRCC) exhibits extensive intratumoral heterogeneity on multiple biological scales, contributing to variable clinical outcomes and limiting the effectiveness of conventional TNM staging, which highlights the urgent need for multiscale integrative analytic frameworks. The lipid-deficient de-clear cell differentiated (DCCD) ccRCC subtype, defined by multi-omics analyses, is associated with adverse outcomes even in early-stage disease. Here, we establish a hierarchical cross-scale framework for the preoperative identification of DCCD-ccRCC. At the highest layer, cross-modal mapping transferred molecular signatures to histological and CT phenotypes, establishing a molecular-to-pathology-to-radiology supervisory bridge. Within this framework, each modality-specific model is designed to mirror the inherent hierarchical structure of tumor biology. PathoDCCD captured multi-scale microscopic features, from cellular morphology and tissue architecture to meso-regional organization. RadioDCCD integrated complementary macroscopic information by combining whole-tumor and its habitat-subregions radiomics with a 2D maximal-section heterogeneity metric. These nested models enabled integrated molecular subtype prediction and clinical risk stratification. Across five cohorts totaling 1,659 patients, PathoDCCD reliably recapitulated molecular subtypes, while RadioDCCD provided reliable preoperative prediction. The consistent predictions identified patients with the poorest clinical outcomes. This cross-scale paradigm unifies molecular biology, computational pathology, and quantitative radiology into a biologically grounded strategy for preoperative noninvasive molecular phenotyping of ccRCC.

---

### Compute the edge p-Laplacian centrality for air traffic network

Authors: Loc Hoang Tran, Bao Nguyen Tran, Luong Anh Tuan Nguyen
Published: 2025-12-13
Categories: cs.SI, cs.LG
arXiv: https://arxiv.org/abs/2512.14749v1

The problem that we would like to solve in this paper is to compute the edge p-Laplacian centrality for the air traffic network. In this problem, instead of computing the edge p-Laplacian centrality directly which is the very hard problem, we convert the air traffic network to the line graph. Finally, we will compute the node p-Laplacian centrality of the line graph which is equivalent to the edge p-Laplacian of the air traffic network. In this paper, the novel un-normalized graph (p-) Laplacian based ranking method will be developed based on the un-normalized graph p-Laplacian operator definitions such as the curvature operator of graph (i.e. the un-normalized graph 1-Laplacian operator) and will be used to compute the node p-Laplacian centrality of the line graph. The results from the experiments show that the un-normalized graph p-Laplacian ranking methods can be implemented successfully.

---

### Composite Classifier-Free Guidance for Multi-Modal Conditioning in Wind Dynamics Super-Resolution

Authors: Jacob Schnell, Aditya Makkar, Gunadi Gani, Aniket Srinivasan Ashok, Darren Lo, Mike Optis, Alexander Wong, Yuhao Chen
Published: 2025-12-13
Categories: cs.LG, cs.AI, cs.CV
arXiv: https://arxiv.org/abs/2512.13729v1

Various weather modelling problems (e.g., weather forecasting, optimizing turbine placements, etc.) require ample access to high-resolution, highly accurate wind data. Acquiring such high-resolution wind data, however, remains a challenging and expensive endeavour. Traditional reconstruction approaches are typically either cost-effective or accurate, but not both. Deep learning methods, including diffusion models, have been proposed to resolve this trade-off by leveraging advances in natural image super-resolution. Wind data, however, is distinct from natural images, and wind super-resolvers often use upwards of 10 input channels, significantly more than the usual 3-channel RGB inputs in natural images. To better leverage a large number of conditioning variables in diffusion models, we present a generalization of classifier-free guidance (CFG) to multiple conditioning inputs. Our novel composite classifier-free guidance (CCFG) can be dropped into any pre-trained diffusion model trained with standard CFG dropout. We demonstrate that CCFG outputs are higher-fidelity than those from CFG on wind super-resolution tasks. We present WindDM, a diffusion model trained for industrial-scale wind dynamics reconstruction and leveraging CCFG. WindDM achieves state-of-the-art reconstruction quality among deep learning models and costs up to $1000\times$ less than classical methods.

---

### CurvaDion: Curvature-Adaptive Distributed Orthonormalization

Authors: Bhavesh Kumar, Roger Jin, Jeffrey Quesnelle
Published: 2025-12-13
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2512.13728v1

As language models scale to trillions of parameters, distributed training across many GPUs becomes essential, yet gradient synchronization over high-bandwidth, low-latency networks remains a critical bottleneck. While recent methods like Dion reduce per-step communication through low-rank updates, they synchronize at every step regardless of the optimization landscape. We observe that synchronization requirements vary dramatically throughout training: workers naturally compute similar gradients in flat regions, making frequent synchronization redundant, while high-curvature regions require coordination to prevent divergence. We introduce CurvaDion, which uses Relative Maximum Momentum Change (RMMC) to detect high-curvature regions requiring synchronization. RMMC leverages momentum dynamics which are already computed during optimization as a computationally tractable proxy for directional curvature, adding only $\mathcal{O}(d)$ operations per layer. We establish theoretical connections between RMMC and loss curvature and demonstrate that CurvaDion achieves 99\% communication reduction while matching baseline convergence across models from 160M to 1.3B parameters.

---

### RAST-MoE-RL: A Regime-Aware Spatio-Temporal MoE Framework for Deep Reinforcement Learning in Ride-Hailing

Authors: Yuhan Tang, Kangxin Cui, Jung Ho Park, Yibo Zhao, Xuan Jiang, Haoze He, Jiangbo Yu, Haris Koutsopoulos, Jinhua Zhao
Published: 2025-12-13
Categories: cs.LG
arXiv: https://arxiv.org/abs/2512.13727v2

Ride-hailing platforms face the challenge of balancing passenger waiting times with overall system efficiency under highly uncertain supply-demand conditions. Adaptive delayed matching, which controls the holding intervals for batched sets of requests and vehicles, reveals an inherent trade-off between matching and pickup delays. The resulting environment with temporally varying request arrival patterns and dynamic congestion calls for more expressive networks with sufficient capacity to capture their non-stationarity. To address the limitations of existing methods that rely on shallow encoders that cannot capture dynamic supply-demand patterns and congestion effects, we introduce the Regime-Aware Spatio-Temporal Mixture-of-Experts (RAST-MoE) framework, which formalizes adaptive delayed matching as a regime-aware Markov Decision Process and equips RL agents with a self-attention MoE encoder. Instead of relying on a single monolithic network, our design allows different experts to specialize automatically in varying operational conditions, improving representation capacity while maintaining per-sample computation efficiency. Despite its modest size of only 12M parameters, our framework consistently outperforms strong baselines. On real-world Uber trajectory data from San Francisco, it reduces average matching delay by 10%, and pickup delay by 15%. In addition, it demonstrates robustness to unseen demand regimes, stable training behavior without reward hacking, and expert specialization to different regimes. This study shows the strength of MoE-enhanced RL for large-scale decision-making tasks with complex spatiotemporal dynamics.

---

### Time-Constrained Recommendations: Reinforcement Learning Strategies for E-Commerce

Authors: Sayak Chakrabarty, Souradip Pal
Published: 2025-12-13
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2512.13726v1

Unlike traditional recommendation tasks, finite user time budgets introduce a critical resource constraint, requiring the recommender system to balance item relevance and evaluation cost. For example, in a mobile shopping interface, users interact with recommendations by scrolling, where each scroll triggers a list of items called slate. Users incur an evaluation cost - time spent assessing item features before deciding to click. Highly relevant items having higher evaluation costs may not fit within the user's time budget, affecting engagement. In this position paper, our objective is to evaluate reinforcement learning algorithms that learn patterns in user preferences and time budgets simultaneously, crafting recommendations with higher engagement potential under resource constraints. Our experiments explore the use of reinforcement learning to recommend items for users using Alibaba's Personalized Re-ranking dataset supporting slate optimization in e-commerce contexts. Our contributions include (i) a unified formulation of time-constrained slate recommendation modeled as Markov Decision Processes (MDPs) with budget-aware utilities; (ii) a simulation framework to study policy behavior on re-ranking data; and (iii) empirical evidence that on-policy and off-policy control can improve performance under tight time budgets than traditional contextual bandit-based methods.

---

### Compressed Causal Reasoning: Quantization and GraphRAG Effects on Interventional and Counterfactual Accuracy

Authors: Steve Nwaiwu, Nipat Jongsawat, Anucha Tungkasthan
Published: 2025-12-13
Categories: cs.AI
arXiv: https://arxiv.org/abs/2512.13725v2

Causal reasoning in Large Language Models spanning association, intervention, and counterfactual inference is essential for reliable decision making in high stakes settings. As deployment shifts toward edge and resource constrained environments, quantized models such as INT8 and NF4 are becoming standard. Yet the impact of precision reduction on formal causal reasoning is poorly understood. To our knowledge, this is the first study to systematically evaluate quantization effects across all three levels of Pearls Causal Ladder. Using a 3000 sample stratified CLadder benchmark, we find that rung level accuracy in Llama 3 8B remains broadly stable under quantization, with NF4 showing less than one percent overall degradation. Interventional queries at rung 2 are the most sensitive to precision loss, whereas counterfactual reasoning at rung 3 is comparatively stable but exhibits heterogeneous weaknesses across query types such as collider bias and backdoor adjustment. Experiments on the CRASS benchmark show near identical performance across precisions, indicating that existing commonsense counterfactual datasets lack the structural sensitivity needed to reveal quantization induced reasoning drift. We further evaluate Graph Retrieval Augmented Generation using ground truth causal graphs and observe a consistent improvement in NF4 interventional accuracy of plus 1.7 percent, partially offsetting compression related degradation. These results suggest that causal reasoning is unexpectedly robust to four bit quantization, graph structured augmentation can selectively reinforce interventional reasoning, and current counterfactual benchmarks fail to capture deeper causal brittleness. This work provides an initial empirical map of compressed causal reasoning and practical guidance for deploying efficient and structurally supported causal AI systems.

---

### Graph AI generates neurological hypotheses validated in molecular, organoid, and clinical systems

Authors: Ayush Noori, Joaquín Polonuer, Katharina Meyer, Bogdan Budnik, Shad Morton, Xinyuan Wang, Sumaiya Nazeen, Yingnan He, Iñaki Arango, Lucas Vittor, Matthew Woodworth, Richard C. Krolewski et al.
Published: 2025-12-13
Categories: q-bio.QM, cs.AI, q-bio.NC
arXiv: https://arxiv.org/abs/2512.13724v1

Neurological diseases are the leading global cause of disability, yet most lack disease-modifying treatments. We present PROTON, a heterogeneous graph transformer that generates testable hypotheses across molecular, organoid, and clinical systems. To evaluate PROTON, we apply it to Parkinson's disease (PD), bipolar disorder (BD), and Alzheimer's disease (AD). In PD, PROTON linked genetic risk loci to genes essential for dopaminergic neuron survival and predicted pesticides toxic to patient-derived neurons, including the insecticide endosulfan, which ranked within the top 1.29% of predictions. In silico screens performed by PROTON reproduced six genome-wide $α$-synuclein experiments, including a split-ubiquitin yeast two-hybrid system (normalized enrichment score (NES) = 2.30, FDR-adjusted $p < 1 \times 10^{-4}$), an ascorbate peroxidase proximity labeling assay (NES = 2.16, FDR $< 1 \times 10^{-4}$), and a high-depth targeted exome sequencing study in 496 synucleinopathy patients (NES = 2.13, FDR $< 1 \times 10^{-4}$). In BD, PROTON predicted calcitriol as a candidate drug that reversed proteomic alterations observed in cortical organoids derived from BD patients. In AD, we evaluated PROTON predictions in health records from $n = 610,524$ patients at Mass General Brigham, confirming that five PROTON-predicted drugs were associated with reduced seven-year dementia risk (minimum hazard ratio = 0.63, 95% CI: 0.53-0.75, $p < 1 \times 10^{-7}$). PROTON generated neurological hypotheses that were evaluated across molecular, organoid, and clinical systems, defining a path for AI-driven discovery in neurological disease.

---

### Made-in China, Thinking in America:U.S. Values Persist in Chinese LLMs

Authors: David Haslett, Linus Ta-Lun Huang, Leila Khalatbari, Janet Hui-wen Hsiao, Antoni B. Chan
Published: 2025-12-13
Categories: cs.CY, cs.AI
arXiv: https://arxiv.org/abs/2512.13723v1

As large language models increasingly mediate access to information and facilitate decision-making, they are becoming instruments in soft power competitions between global actors such as the United States and China. So far, language models seem to be aligned with the values of Western countries, but evidence for this ethical bias comes mostly from models made by American companies. The current crop of state-of-the-art models includes several made in China, so we conducted the first large-scale investigation of how models made in China and the USA align with people from China and the USA. We elicited responses to the Moral Foundations Questionnaire 2.0 and the World Values Survey from ten Chinese models and ten American models, and we compared their responses to responses from thousands of Chinese and American people. We found that all models respond to both surveys more like American people than like Chinese people. This skew toward American values is only slightly mitigated when prompting the models in Chinese or imposing a Chinese persona on the models. These findings have important implications for a near future in which large language models generate much of the content people consume and shape normative influence in geopolitics.

---

### Advancing Cache-Based Few-Shot Classification via Patch-Driven Relational Gated Graph Attention

Authors: Tasweer Ahmad, Arindam Sikdar, Sandip Pradhan, Ardhendu Behera
Published: 2025-12-13
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.12498v1

Few-shot image classification remains difficult under limited supervision and visual domain shift. Recent cache-based adaptation approaches (e.g., Tip-Adapter) address this challenge to some extent by learning lightweight residual adapters over frozen features, yet they still inherit CLIP's tendency to encode global, general-purpose representations that are not optimally discriminative to adapt the generalist to the specialist's domain in low-data regimes. We address this limitation with a novel patch-driven relational refinement that learns cache adapter weights from intra-image patch dependencies rather than treating an image embedding as a monolithic vector. Specifically, we introduce a relational gated graph attention network that constructs a patch graph and performs edge-aware attention to emphasize informative inter-patch interactions, producing context-enriched patch embeddings. A learnable multi-aggregation pooling then composes these into compact, task-discriminative representations that better align cache keys with the target few-shot classes. Crucially, the proposed graph refinement is used only during training to distil relational structure into the cache, incurring no additional inference cost beyond standard cache lookup. Final predictions are obtained by a residual fusion of cache similarity scores with CLIP zero-shot logits. Extensive evaluations on 11 benchmarks show consistent gains over state-of-the-art CLIP adapter and cache-based baselines while preserving zero-shot efficiency. We further validate battlefield relevance by introducing an Injured vs. Uninjured Soldier dataset for casualty recognition. It is motivated by the operational need to support triage decisions within the "platinum minutes" and the broader "golden hour" window in time-critical UAV-driven search-and-rescue and combat casualty care.

---

### Policy Optimization for Dynamic Heart Transplant Allocation

Authors: Ioannis Anagnostides, Zachary W. Sollie, Arman Kilic, Tuomas Sandholm
Published: 2025-12-13
Categories: cs.LG
arXiv: https://arxiv.org/abs/2512.12497v1

Heart transplantation is a viable path for patients suffering from advanced heart failure, but this lifesaving option is severely limited due to donor shortage. Although the current allocation policy was recently revised in 2018, a major concern is that it does not adequately take into account pretransplant and post-transplant mortality. In this paper, we take an important step toward addressing these deficiencies. To begin with, we use historical data from UNOS to develop a new simulator that enables us to evaluate and compare the performance of different policies. We then leverage our simulator to demonstrate that the status quo policy is considerably inferior to the myopic policy that matches incoming donors to the patient who maximizes the number of years gained by the transplant. Moreover, we develop improved policies that account for the dynamic nature of the allocation process through the use of potentials -- a measure of a patient's utility in future allocations that we introduce. We also show that batching together even a handful of donors -- which is a viable option for a certain type of donors -- further enhances performance. Our simulator also allows us to evaluate the effect of critical, and often unexplored, factors in the allocation, such as geographic proximity and the tendency to reject offers by the transplant centers.

---

### AI-Driven Early Warning Systems for Student Success: Discovering Static Feature Dominance in Temporal Prediction Models

Authors: Vaarunay Kaushal, Rajib Mall
Published: 2025-12-13
Categories: cs.LG
arXiv: https://arxiv.org/abs/2512.12493v1

Early identification of at-risk students is critical for effective intervention in online learning environments. This study extends temporal prediction analysis to Week 20 (50% of course duration), comparing Decision Tree and Long Short- Term Memory (LSTM) models across six temporal snapshots. Our analysis reveals that different performance metrics matter at different intervention stages: high recall is critical for early intervention (Weeks 2-4), while balanced precision-recall is important for mid-course resource allocation (Weeks 8-16), and high precision becomes paramount in later stages (Week 20). We demonstrate that static demographic features dominate predictions (68% importance), enabling assessment-free early prediction. The LSTM model achieves 97% recall at Week 2, making it ideal for early intervention, while Decision Tree provides stable balanced performance (78% accuracy) during mid-course. By Week 20, both models converge to similar recall (68%), but LSTM achieves higher precision (90% vs 86%). Our findings also suggest that model selection should depend on intervention timing, and that early signals (Weeks 2-4) are sufficient for reliable initial prediction using primarily demographic and pre-enrollment information.

---

### Adaptive Detector-Verifier Framework for Zero-Shot Polyp Detection in Open-World Settings

Authors: Shengkai Xu, Hsiang Lun Kao, Tianxiang Xu, Honghui Zhang, Junqiao Wang, Runmeng Ding, Guanyu Liu, Tianyu Shi, Zhenyu Yu, Guofeng Pan, Ziqian Bi, Yuqi Ouyang
Published: 2025-12-13
Categories: cs.CV, cs.CL
arXiv: https://arxiv.org/abs/2512.12492v2

Polyp detectors trained on clean datasets often underperform in real-world endoscopy, where illumination changes, motion blur, and occlusions degrade image quality. Existing approaches struggle with the domain gap between controlled laboratory conditions and clinical practice, where adverse imaging conditions are prevalent. In this work, we propose AdaptiveDetector, a novel two-stage detector-verifier framework comprising a YOLOv11 detector with a vision-language model (VLM) verifier. The detector adaptively adjusts per-frame confidence thresholds under VLM guidance, while the verifier is fine-tuned with Group Relative Policy Optimization (GRPO) using an asymmetric, cost-sensitive reward function specifically designed to discourage missed detections -- a critical clinical requirement. To enable realistic assessment under challenging conditions, we construct a comprehensive synthetic testbed by systematically degrading clean datasets with adverse conditions commonly encountered in clinical practice, providing a rigorous benchmark for zero-shot evaluation. Extensive zero-shot evaluation on synthetically degraded CVC-ClinicDB and Kvasir-SEG images demonstrates that our approach improves recall by 14 to 22 percentage points over YOLO alone, while precision remains within 0.7 points below to 1.7 points above the baseline. This combination of adaptive thresholding and cost-sensitive reinforcement learning achieves clinically aligned, open-world polyp detection with substantially fewer false negatives, thereby reducing the risk of missed precancerous polyps and improving patient outcomes.

---

### GoMS: Graph of Molecule Substructure Network for Molecule Property Prediction

Authors: Shuhui Qu, Cheolwoo Park
Published: 2025-12-13
Categories: cs.LG
arXiv: https://arxiv.org/abs/2512.12489v1

While graph neural networks have shown remarkable success in molecular property prediction, current approaches like the Equivariant Subgraph Aggregation Networks (ESAN) treat molecules as bags of independent substructures, overlooking crucial relationships between these components. We present Graph of Molecule Substructures (GoMS), a novel architecture that explicitly models the interactions and spatial arrangements between molecular substructures. Unlike ESAN's bag-based representation, GoMS constructs a graph where nodes represent subgraphs and edges capture their structural relationships, preserving critical topological information about how substructures are connected and overlap within the molecule. Through extensive experiments on public molecular datasets, we demonstrate that GoMS outperforms ESAN and other baseline methods, with particularly improvements for large molecules containing more than 100 atoms. The performance gap widens as molecular size increases, demonstrating GoMS's effectiveness for modeling industrial-scale molecules. Our theoretical analysis demonstrates that GoMS can distinguish molecules with identical subgraph compositions but different spatial arrangements. Our approach shows particular promise for materials science applications involving complex molecules where properties emerge from the interplay between multiple functional units. By capturing substructure relationships that are lost in bag-based approaches, GoMS represents a significant advance toward scalable and interpretable molecular property prediction for real-world applications.

---

### The American Ghost in the Machine: How language models align culturally and the effects of cultural prompting

Authors: James Luther, Donald Brown
Published: 2025-12-13
Categories: cs.CL
arXiv: https://arxiv.org/abs/2512.12488v1

Culture is the bedrock of human interaction; it dictates how we perceive and respond to everyday interactions. As the field of human-computer interaction grows via the rise of generative Large Language Models (LLMs), the cultural alignment of these models become an important field of study. This work, using the VSM13 International Survey and Hofstede's cultural dimensions, identifies the cultural alignment of popular LLMs (DeepSeek-V3, V3.1, GPT-5, GPT-4.1, GPT-4, Claude Opus 4, Llama 3.1, and Mistral Large). We then use cultural prompting, or using system prompts to shift the cultural alignment of a model to a desired country, to test the adaptability of these models to other cultures, namely China, France, India, Iran, Japan, and the United States. We find that the majority of the eight LLMs tested favor the United States when the culture is not specified, with varying results when prompted for other cultures. When using cultural prompting, seven of the eight models shifted closer to the expected culture. We find that models had trouble aligning with Japan and China, despite two of the models tested originating with the Chinese company DeepSeek.

---

### More Than the Final Answer: Improving Visual Extraction and Logical Consistency in Vision-Language Models

Authors: Hoang Anh Just, Yifei Fan, Handong Zhao, Jiuxiang Gu, Ruiyi Zhang, Simon Jenni, Kushal Kafle, Ruoxi Jia, Jing Shi
Published: 2025-12-13
Categories: cs.CV
arXiv: https://arxiv.org/abs/2512.12487v1

Reinforcement learning from verifiable rewards (RLVR) has recently been extended from text-only LLMs to vision-language models (VLMs) to elicit long-chain multimodal reasoning. However, RLVR-trained VLMs still exhibit two persistent failure modes: inaccurate visual extraction (missing or hallucinating details) and logically inconsistent chains-of-thought, largely because verifiable signals supervise only the final answer. We propose PeRL-VL (Perception and Reasoning Learning for Vision-Language Models), a decoupled framework that separately improves visual perception and textual reasoning on top of RLVR. For perception, PeRL-VL introduces a VLM-based description reward that scores the model's self-generated image descriptions for faithfulness and sufficiency. For reasoning, PeRL-VL adds a text-only Reasoning SFT stage on logic-rich chain-of-thought data, enhancing coherence and logical consistency independently of vision. Across diverse multimodal benchmarks, PeRL-VL improves average Pass@1 accuracy from 63.3% (base Qwen2.5-VL-7B) to 68.8%, outperforming standard RLVR, text-only reasoning SFT, and naive multimodal distillation from GPT-4o.

---

### Mage: Cracking Elliptic Curve Cryptography with Cross-Axis Transformers

Authors: Lily Erickson
Published: 2025-12-13
Categories: cs.CR, cs.AI
arXiv: https://arxiv.org/abs/2512.12483v3

With the advent of machine learning and quantum computing, the 21st century has gone from a place of relative algorithmic security, to one of speculative unease and possibly, cyber catastrophe. Modern algorithms like Elliptic Curve Cryptography (ECC) are the bastion of current cryptographic security protocols that form the backbone of consumer protection ranging from Hypertext Transfer Protocol Secure (HTTPS) in the modern internet browser, to cryptographic financial instruments like Bitcoin. And there's been very little work put into testing the strength of these ciphers. Practically the only study that I could find was on side-channel recognition, a joint paper from the University of Milan, Italy and King's College, London\cite{battistello2025ecc}. These algorithms are already considered bulletproof by many consumers, but exploits already exist for them, and with computing power and distributed, federated compute on the rise, it's only a matter of time before these current bastions fade away into obscurity, and it's on all of us to stand up when we notice something is amiss, lest we see such passages claim victims in that process. In this paper, we seek to explore the use of modern language model architecture in cracking the association between a known public key, and its associated private key, by intuitively learning to reverse engineer the public keypair generation process, effectively solving the curve. Additonally, we attempt to ascertain modern machine learning's ability to memorize public-private secp256r1 keypairs, and to then test their ability to reverse engineer the public keypair generation process. It is my belief that proof-for would be equally valuable as proof-against in either of these categories. Finally, we'll conclude with some number crunching on where we see this particular field heading in the future.

---