Large Language Models (LLMs) are increasingly applied to domains that require reasoning about other agents' behavior, such as negotiation, policy design, and market simulation, yet existing research has mostly evaluated their adherence to equilibrium play or their exhibited depth of reasoning. Whether they display genuine strategic thinking, understood as the coherent formation of beliefs about other agents, evaluation of possible actions, and choice based on those beliefs, remains unexplored. We develop a framework to identify this ability by disentangling beliefs, evaluation, and choice in static, complete-information games, and apply it across a series of non-cooperative environments. By jointly analyzing models' revealed choices and reasoning traces, and introducing a new context-free game to rule out imitation from memorization, we show that current frontier models exhibit belief-coherent best-response behavior at targeted reasoning depths. When unconstrained, they self-limit their depth of reasoning and form differentiated conjectures about human and synthetic opponents, revealing an emergent form of meta-reasoning. Under increasing complexity, explicit recursion gives way to internally generated heuristic rules of choice that are stable, model-specific, and distinct from known human biases. These findings indicate that belief coherence, meta-reasoning, and novel heuristic formation can emerge jointly from language modeling objectives, providing a structured basis for the study of strategic cognition in artificial agents.

Robotic tactile perception is a complex process involving several computational steps performed at different levels. Tactile information is shaped by the interplay of robot actions, the mechanical properties of its body, and the software that processes the data. In this respect, high-level computation, required to process and extract information, is commonly performed by adapting existing techniques from other domains, such as computer vision, which expects input data to be properly structured. Therefore, it is necessary to transform tactile sensor data to match a specific data structure. This operation directly affects the tactile information encoded and, as a consequence, the task execution. This survey aims to address this specific aspect of the tactile perception pipeline, namely Data Representation. The paper first clearly defines its contributions to the perception pipeline and then reviews how previous studies have dealt with the problem of representing tactile information, investigating the relationships among hardware, representations, and high-level computation methods. The analysis has led to the identification of six structures commonly used in the literature to represent data. The manuscript provides discussions and guidelines for properly selecting a representation depending on operating conditions, including the available hardware, the tactile information required to be encoded, and the task at hand.

This work addresses the challenge of using a deep learning model to prune graphs and the ability of this method to integrate explainability into spatio-temporal problems through a new approach. Instead of applying explainability to the model's behavior, we seek to gain a better understanding of the problem itself. To this end, we propose a novel model that integrates an optimized pruning mechanism capable of removing nodes from the graph during the training process, rather than doing so as a separate procedure. This integration allows the architecture to learn how to minimize prediction error while selecting the most relevant nodes. Thus, during training, the model searches for the most relevant subset of nodes, obtaining the most important elements of the problem, facilitating its analysis. To evaluate the proposed approach, we used several widely used traffic datasets, comparing the accuracy obtained by pruning with the model and with other methods. The experiments demonstrate that our method is capable of retaining a greater amount of information as the graph reduces in size compared to the other methods used. These results highlight the potential of pruning as a tool for developing models capable of simplifying spatio-temporal problems, thereby obtaining their most important elements.

Key Points: We compare linear regression, LSTM, and Transformer models for predicting terrestrial water storage at basin scale over the globe. Linear regression remains a robust benchmark, outperforming LSTM and Transformer models in various tasks. Traditional statistical models and global datasets that capture human and natural impacts are essential for deep learning model evaluation. 2 Abstract Recent advances in machine learning such as Long Short - Term Memory (LSTM) models and Transformers have been widely adopted in hydrological applications, demonstrating impressive performance amongst deep learning models and outperforming physical models in various tasks. However, their superiority in predicting land surface states such as terrestrial water storage (TWS) that are dominated by many factors such as natural variability and human driven modifications remains unclear. Here, using the open - access, globally representative HydroGlobe dataset - comprising a baseline version derived solely from a land surface model simulation and an advanced version incorporating multi - source remote sensing data assimilation - we show that linear regres sion is a robust benchmark, outperforming the more complex LSTM and Temporal Fusion Transformer for TWS prediction. Our findings highlight the importance of including traditional statistical models as benchmarks when developing and evaluating deep learning models. Additionally, we emphasize the critical need to establish globally representative benchmark datasets that capture the combined impact of natural variability and human interventions. Plain Language Summary Recent progress in machine learning has led to the widespread use of deep learning models in studying land freshwater systems, but it remains uncertain if they're always the best tools for such applications . In this study, we use a new, global dataset called HydroGlobe to test different data - driven models. Surprisingly, we find that a basic linear regression model -- one of the simplest tools -- actually performs better than more complex models like LSTM and Transformers in predicting land water storage. Our resu lts suggest that researchers should always compare deep learning models against simpler traditional statistical benchmarks, and that having high - quality, global datasets that include both natural and human effects is crucial for building better deep learning models. 1 Introduction Terrestrial water storage (TWS) is a key indicator of the world's freshwater availability, encompassing all forms of water stored on and beneath the land surface, including soil moisture, groundwater, surface water, and snow. As a fundamental component of the global hydrological cycle, accurate TWS estimates are essential for applications related to preserving ecosystems, supporting agriculture, and ensuring water and food security for livelihoods.

Today's deep learning architectures are primarily based on perceptron models, which do not capture the oscillatory dynamics characteristic of biological neurons. Although oscillatory systems have recently gained attention for their closer resemblance to neural behavior, they still fall short of modeling the intricate spatio-temporal interactions observed in natural neural circuits. In this paper, we propose a bio-inspired oscillatory state system (BioOSS) designed to emulate the wave-like propagation dynamics critical to neural processing, particularly in the prefrontal cortex (PFC), where complex activity patterns emerge. BioOSS comprises two interacting populations of neurons: p neurons, which represent simplified membrane-potential-like units inspired by pyramidal cells in cortical columns, and o neurons, which govern propagation velocities and modulate the lateral spread of activity. Through local interactions, these neurons produce wave-like propagation patterns. The model incorporates trainable parameters for damping and propagation speed, enabling flexible adaptation to task-specific spatio-temporal structures. We evaluate BioOSS on both synthetic and real-world tasks, demonstrating superior performance and enhanced interpretability compared to alternative architectures.

The performance gains of LLMs have historically been driven by scaling up model size and training data. However, the rapidly diminishing availability of high-quality training data is introducing a fundamental bottleneck, shifting the focus of research toward inference-time scaling. This paradigm uses additional computation at the time of deployment to substantially improve LLM performance on downstream tasks without costly model re-training. This review systematically surveys the diverse techniques contributing to this new era of inference-time scaling, organizing the rapidly evolving field into two comprehensive perspectives: Output-focused and Input-focused methods. Output-focused techniques encompass complex, multi-step generation strategies, including reasoning (e.g., CoT, ToT, ReAct), various search and decoding methods (e.g., MCTS, beam search), training for long CoT (e.g., RLVR, GRPO), and model ensemble methods. Input-focused techniques are primarily categorized by few-shot and RAG, with RAG as the central focus. The RAG section is further detailed through a structured examination of query expansion, data, retrieval and reranker, LLM generation methods, and multi-modal RAG.

In this paper, we propose a novel framework, Disentangled Style-Content GAN (DISC-GAN), which integrates style-content disentanglement with a cluster-specific training strategy towards photorealistic underwater image synthesis. The quality of synthetic underwater images is challenged by optical distortions due to phenomena such as color attenuation and turbidity. These phenomena are represented by distinct stylistic variations across different wa-terbodies, such as changes in tint and haze. While generative models are well-suited to capture complex patterns, they often lack the ability to model the non-uniform stylistic conditions of diverse underwater environments. T o address these challenges, we employ K-means clustering to partition a dataset into style-specific domains. W e use separate encoders to get latent spaces for style and content; we further integrate these latent representations via Adaptive Instance Normalization (AdaIN) and decode the result to produce the final synthetic image. The model is trained independently on each style cluster to preserve domain-specific characteristics. Our framework demonstrates state-of-the-art performance, obtaining a Structural Similarity Index (SSIM) of 0.9012, an average Peak Signal-to-Noise Ratio (PSNR) of 32.5118 dB, and a Fr echet Inception Distance (FID) of 13.3728.

The language of Hiligaynon, spoken predominantly by the people of Panay Island, Negros Occidental, and Soccsksargen in the Philippines, remains underrepresented in language processing research due to the absence of annotated corpora and baseline models. This study introduces HiligayNER, the first publicly available baseline model for the task of Named Entity Recognition (NER) in Hiligaynon. The dataset used to build HiligayNER contains over 8,000 annotated sentences collected from publicly available news articles, social media posts, and literary texts. Two Transformer-based models, mBERT and XLM-RoBERTa, were fine-tuned on this collected corpus to build versions of HiligayNER. Evaluation results show strong performance, with both models achieving over 80% in precision, recall, and F1-score across entity types. Furthermore, cross-lingual evaluation with Cebuano and Tagalog demonstrates promising transferability, suggesting the broader applicability of HiligayNER for multilingual NLP in low-resource settings. This work aims to contribute to language technology development for underrepresented Philippine languages, specifically for Hiligaynon, and support future research in regional language processing.

In financial markets, Graph Neural Networks have been successfully applied to modeling relational data, effectively capturing nonlinear inter-stock dependencies. Yet, existing models often fail to efficiently propagate messages during macroeconomic shocks. In this paper, we propose OmniGNN, an attention-based multi-relational dynamic GNN that integrates macroeconomic context via heterogeneous node and edge types for robust message passing. Central to OmniGNN is a sector node acting as a global intermediary, enabling rapid shock propagation across the graph without relying on long-range multi-hop diffusion. The model leverages Graph Attention Networks (GAT) to weigh neighbor contributions and employs Transformers to capture temporal dynamics across multiplex relations. Experiments show that OmniGNN outperforms existing stock prediction models on public datasets, particularly demonstrating strong robustness during the COVID-19 period.

Abstract-- As autonomous vehicles become an essential component of modern transportation, they are increasingly vulnerable to threats such as GPS spoofing attacks. This study presents an adaptive detection approach utilizing a dynamically tuned Density Based Spatial Clustering of Applications with Noise (DBSCAN) algorithm, designed to adjust the detection threshold (ε) in real-time. The threshold is updated based on the recursive mean and standard deviation of displacement errors between GPS and in-vehicle sensors data, but only at instances classified as non-anomalous. Furthermore, an initial threshold, determined from 120,000 clean data samples, ensures the capability to identify even subtle and gradual GPS spoofing attempts from the beginning. T o assess the performance of the proposed method, five different subsets from the real-world Honda Research Institute Driving Dataset (HDD) are selected to simulate both large and small magnitude GPS spoofing attacks. The modified algorithm effectively identifies turn-by-turn, stop, overshoot, and multiple small biased spoofing attacks, achieving detection accuracies of 98.62 1%, 99.96 0.1%, 99.88 0.1%, and 98.38 0.1%, respectively. This work provides a substantial advancement in enhancing the security and safety of A Vs against GPS spoofing threats.