Vision Transformer (ViT) has become one of the cornerstones of the computer vision field, demonstrating exceptional performance. However, its inherent high computational complexity and inference latency still pose significant obstacles for deployment in resource-constrained environments. Token pruning, by removing less informative tokens, offers an effective strategy to reduce computational overhead. However, existing pruning methods largely rely on static or local token importance scores. This myopic approach fundamentally overlooks the sequential dependency of pruning decisions and fails to capture the interaction effects between pruning decisions across layers, often neglecting the global interactions between mask variables. To address this limitation, we propose V-Pruner, a fast and globally-informed token pruning framework for Vision Transformer. V-Pruner first leverages Fisher information to perform an initial assessment of token importance, providing a principled initial prior for pruning decisions. Building on this, V-Pruner introduces a Reinforcement Learning (RL) Proximal Policy Optimization (PPO) algorithm, refining token pruning into a global sequential decision process. The algorithm combines a composite reward signal that incorporates both model performance and computational cost to guide policy exploration, effectively evaluating the long-term impact of different pruning decision combinations on global model performance. Extensive experiments on ViT-L, DeiT-B, DeiT-S, and DeiT-T demonstrate that V-Pruner achieves a better balance between accuracy, GFLOPs, inference speed, and training time, surpassing existing mainstream ViT pruning algorithms in overall performance.

Vision-language models are increasingly deployed as computer-use agents (CUAs) that operate desktops and browsers.Top-performing CUAs are framework-based systems that decompose planning and execution, while endto-end screenshot-to-action policies are easier to deploy but lag behind on benchmarks such as OSWorld-Verified.GUI datasets like OSWorld pose two bottlenecks: they expose only a few hundred interactive, verifiable tasks and environments, and expert trajectories must be gathered by interacting with these environments, making such data hard to scale.We therefore ask how reinforcement learning from verifiable rewards (RLVR) can best exploit a small pool of exist expert trajectories to train end-to-end policies.Naïvely mixing these offpolicy traces into on-policy RLVR is brittle: even after format conversion, expert trajectories exhibit structural mismatch and distribution shift from the learner.We propose BEPA (Bi-Level Expert-to-Policy Assimilation), which turns static expert traces into policy-aligned guidance via self-rolled reachable trajectories under the base policy (LEVEL-1) and a pertask, dynamically updated cache used in RLVR (LEVEL-2).On OSWorld-Verified, BEPA improves UITARS1.5-7Bsuccess from 22.87% to 32.13% and raises a held-out split from 5.74% to 10.30%, with consistent gains on MMBench-GUI and Online-Mind2Web.Our code and data are available at

GUI agents that interact with graphical interfaces on behalf of users are a promising direction for practical AI assistants, yet training them is hindered by scarce suitable environments. We present InfiniteWeb, a system that automatically generates functional web environments at scale for GUI agent training. While LLMs perform well on generating a single webpage, building a realistic and functional website with many interconnected pages faces challenges. We address these challenges through unified specification, task-centric test-driven development, and combining website seed variation with reference design images. Our system also generates verifiable task evaluators enabling dense reward signals for reinforcement learning. Experiments show that our system surpasses commercial coding agents at realistic website construction, and GUI agents trained on our generated environments achieve significant performance improvements on OSWorld and Online-Mind2Web, demonstrating the effectiveness of the proposed system.

In recent years, Vision Transformers (ViT) have achieved significant success in various complex visual tasks, but they also come with substantial computational costs and memory overheads. To address this issue, lightweight Vision Transformers have become an important research direction. Current research on lightweight ViTs mainly focuses on combining CNNs and Transformers, leveraging the advantages of CNNs in local feature extraction while utilizing Transformers’ ability to model global context. However, existing lightweight models often suffer from an imbalance in processing low-frequency global information and high-frequency local information. While sparse attention mechanisms effectively capture global context and reduce computational load, they typically adopt relatively simple strategies for handling high-frequency local information, failing to fully exploit the details of local features. To address this issue, we introduce a new lightweight Vision Transformer model, A Lightweight Vision Transformer for Balancing Global and Local Information (GLFormer). GLFormer combines dynamic weight adjustment with context-aware mechanisms to effectively aggregate high-frequency local information, optimizing the balance between global and local information. Additionally, we introduce a Depth Perception Feed-Forward Network (DPFFN), which further enhances feature fusion and detail refinement, thus enhancing the model’s performance and its capacity to generalize. Based on GLFormer and DPFFN, we design a novel visual backbone network—GLNet. Extensive experimental results show that GLNet consistently demonstrates excellent performance across various tasks, while maintaining a relatively low computational cost.

Large language models (LLMs) have achieved significant success in complex tasks across various domains, but these achievements come with high computational costs and long inference delays. Pruning, as an effective optimization technique, simplifies model structures by removing redundant components, thereby improving model generalization and operational efficiency. Although existing pruning retraining-free algorithms perform excellently in pruning time, these algorithms often focus on local optimal solutions in encoder-based language models, lacking comprehensive exploration of global optimal solutions, which may affect the overall model performance. To address this issue, we propose a novel retraining-free structured pruning algorithm, named RL-Pruner. The algorithm consists of two main stages: the Mask Rearrangement Based on Asynchronous Advantage Actor-Critic (MA3C) stage and the BiConjugate Gradient Solver for Mask Tuning (BGMT) stage. It aims to explore the intra-layer interactions of mask variables and efficiently find the global optimal solution without requiring retraining. We evaluate this method using BERTBASEand DistilBERT models on the GLUE and SQuAD benchmark tests. Experimental results show that RL-Pruner significantly improves accuracy on the SQuAD1.1benchmark. Under a 60% FLOPs constraint, compared with existing pruning retraining-free algorithms, the F1 score increases by 4.25%.

With the explosive growth of data volume, distributed machine learning has become the mainstream approach for training deep neural networks. However, distributed machine learning incurs non-trivial communication overhead. To this end, various compression schemes are proposed to alleviate the communication volume among nodes. Nevertheless, existing compression schemes, such as gradient quantization or gradient sparsification, suffer from low compression ratios and/or high computational overheads. Recent studies advocate leveraging sketch techniques to assist these schemes. However, the limitations of gradient quantization and gradient sparsification remain. In this paper, we propose SK-Gradient, a novel gradient compression scheme that solely builds on sketch. The core component of SK-Gradient is a novel sketch namely FGC Sketch that is tailored for gradient compression. FGC Sketch precomputes the costly hash functions to alleviate computational overheads. Its simplified design makes it convenient for GPU acceleration. In addition, SK-Gradient leverages various techniques including selective gradient compression and periodic synchronization strategy to improve computational efficiency and compression accuracy. Compared with the state-of-the-art schemes, SK-Gradient achieves up to 92.9% reduction in computational overhead and up to 95.2% improvement in training speedups at the same compression ratio.

Automated driving object detection has always been a challenging task in computer vision due to environmental uncertainties. These uncertainties include significant differences in object sizes and encountering the class unseen. It may result in poor performance when traditional object detection models are directly applied to automated driving detection. Because they usually presume fixed categories of common traffic participants, such as pedestrians and cars. Worsely, the huge class imbalance between common and novel classes further exacerbates performance degradation. To address the issues stated, we propose OpenNet to moderate the class imbalance with the Balanced Loss, which is based on Cross Entropy Loss. Besides, we adopt an inductive layer based on gradient reshaping to fast learn new classes with limited samples during incremental learning. To against catastrophic forgetting, we employ normalized feature distillation. By the way, we improve multi-scale detection robustness and unknown class recognition through FPN and energy-based detection, respectively. The Experimental results upon the CODA dataset show that the proposed method can obtain better performance than that of the existing methods.

Understanding and improving global mobility has gained increased interest during the last decades. However, studies on the railway network are spatially limited so far, mostly investigating the domestic network of a country. Data availability is a major limiting factor for the analysis of these networks. Despite the increased open data movement, network operators are often reluctant to publish their infrastructure and passenger data. Existing large-scale studies usually make use of hand-collected data, for instance, based on historical cartographies. In this paper, we develop and implement a methodology to extract the worldwide railway skeleton network from the open data repository OpenStreetMap, where nodes are stations/waypoints and links are weights with information such as spatial distance, gauge, and maximum speed. We describe how we solved several data cleansing and scalability issues and developed network simplification techniques, in order to obtain an adequate representation of the network. We show that the network breaks down into few large and many small components. Furthermore, we show that this public data set can be used for efficient minimum travel time estimation between stations or cities. This paper leads to the development of a new research data set and contributes toward the ability of analyzing global mobility patterns, particularly regarding multimodality and cross-country transportation.