March 2026

arXiv:2603.17063v1 Announce Type: new Abstract: Transformers are the dominant architecture in AI, yet why they work remains poorly understood. This paper offers a precise answer: a transformer is a Bayesian network. We establish this in five ways. First, we prove that every sigmoid transformer with any weights implements weighted loopy belief propagation on its implicit factor graph. One layer is one round of BP. This holds for any weights — trained, random, or constructed. Formally verified against standard […]

Ver mais

arXiv:2603.17061v1 Announce Type: new Abstract: Collecting everyday speech data for prosodic analysis is challenging due to the confounding of prosody and semantics, privacy constraints, and participant compliance. We introduce and empirically evaluate a content-controlled, privacy-first smartphone protocol that uses scripted read-aloud sentences to standardize lexical content (including prompt valence) while capturing natural variation in prosodic delivery. The protocol performs on-device prosodic feature extraction, deletes raw audio immediately, and transmits only derived features for analysis. We deployed the protocol […]

Ver mais

arXiv:2603.17058v1 Announce Type: new Abstract: Nash equilibria provide a principled framework for modeling interactions in multi-agent decision-making and control. However, many equilibrium-seeking methods implicitly assume that each agent has access to the other agents’ objectives and constraints, an assumption that is often unrealistic in practice. This letter studies a class of asymmetric-information two-player constrained games with decoupled feasible sets, in which Player 1 knows its own objective and constraints while Player 2 is available only through a best-response […]

Ver mais

arXiv:2603.17056v1 Announce Type: new Abstract: Reliable terrain perception is a fundamental requirement for autonomous navigation in unstructured, off-road environments. Desert landscapes present unique challenges due to low chromatic contrast between terrain categories, extreme lighting variability, and sparse vegetation that defy the assumptions of standard road-scene segmentation models. We present DesertFormer, a semantic segmentation pipeline for off-road desert terrain analysis based on SegFormer B2 with a hierarchical Mix Transformer (MiT-B2) backbone. The system classifies terrain into ten ecologically meaningful […]

Ver mais

arXiv:2603.17055v1 Announce Type: new Abstract: Image Restoration (IR) agents, leveraging multimodal large language models to perceive degradation and invoke restoration tools, have shown promise in automating IR tasks. However, existing IR agents typically lack an insight summarization mechanism for past interactions, which results in an exhaustive search for the optimal IR tool. To address this limitation, we propose a portrait-aware IR agent, dubbed PaAgent, which incorporates a self-evolving portrait bank for IR tools and Retrieval-Augmented Generation (RAG) to […]

Ver mais

arXiv:2603.17054v1 Announce Type: new Abstract: Reliable and resilient communication is essential for disaster recovery and emergency response, yet terrestrial infrastructure often fails during large-scale natural disasters. This paper proposes a High-Altitude Platform Station (HAPS) and Reconfigurable Intelligent Surfaces (RIS)-assisted Internet of Things (IoT) communication system to restore connectivity in disaster-affected areas. Distributed IoT sensors collect critical environmental data and forward it to nearby gateways via short-range links, while the HAPS-RIS system provides backhaul to these gateways. To overcome […]

Ver mais

arXiv:2603.17052v1 Announce Type: new Abstract: Vector quantization is a technique in machine learning that discretizes continuous representations into a set of discrete vectors. It is widely employed in tokenizing data representations for large language models, diffusion models, and other generative models. Despite its prevalence, the characteristics and behaviors of vector quantization in generative models remain largely underexplored. In this study, we systematically investigate the issue of collapses in vector quantization, where collapsed representations are observed across discrete codebook […]

Ver mais

arXiv:2603.17051v1 Announce Type: new Abstract: Distilled autoregressive (AR) video models enable efficient streaming generation but frequently misalign with human visual preferences. Existing reinforcement learning (RL) frameworks are not naturally suited to these architectures, typically requiring either expensive re-distillation or solver-coupled reverse-process optimization that introduces considerable memory and computational overhead. We present Astrolabe, an efficient online RL framework tailored for distilled AR models. To overcome existing bottlenecks, we introduce a forward-process RL formulation based on negative-aware fine-tuning. By contrasting […]

Ver mais

arXiv:2603.17048v1 Announce Type: new Abstract: Modern neural networks achieve strong performance but remain difficult to interpret in high-dimensional visual domains. Counterfactual explanations (CFEs) provide a principled approach to interpreting black-box predictions by identifying minimal input changes that alter model outputs. However, existing CFE methods often rely on dataset-specific generative models and incur substantial computational cost, limiting their scalability to high-resolution data. We propose SCE-LITE-HQ, a scalable framework for counterfactual generation that leverages pretrained generative foundation models without task-specific […]

Ver mais

arXiv:2603.17047v1 Announce Type: new Abstract: We study $(alpha,beta)$-spanners for weighted graphs. We propose a simple greedy completion procedure which starts from a sparse initial graph, and repeatedly fixes pairs of vertices with a bad stretch, generalizing Kunedsen’s additive completion [SWAT ’14]. As an application, we construct $(k,k-1)$-spanners for weighted graphs of size $tilde{O}(n^{1+1/k})$, which were previously unknown.

Ver mais