arXiv subjects

Yue Wang

Publications and source records attributed to Yue Wang.

At least 235 records · Page 13Linked to original sources

arXiv2026

InfiniDepth: Arbitrary-Resolution and Fine-Grained Depth Estimation with Neural Implicit Fields

Existing depth estimation methods are fundamentally limited to predicting depth on discrete image grids. Such representations restrict their scalability to arbitrary output resolutions and hinder the geometric detail recovery. This paper introduces InfiniDepth, which represents depth as neural implicit fields. Through a simple yet effective local implicit decoder, we can query depth at continuous 2D coordinates, enabling arbitrary-resolution and fine-grained depth estimation. To better assess our method's capabilities, we curate a high-quality 4K synthetic benchmark from five different games, spanning diverse scenes with rich geometric and appearance details. Extensive experiments demonstrate that InfiniDepth achieves state-of-the-art performance on both synthetic and real-world benchmarks across relative and metric depth estimation tasks, particularly excelling in fine-detail regions. It also benefits the task of novel view synthesis under large viewpoint shifts, producing high-quality results with fewer holes and artifacts.

cs.CV↗

arXiv2026

On the Ultra-Long Gamma-Ray Transient GRB 250702B/EP250702

GRB 250702B/EP250702a is an interesting long-duration gamma-ray transient whose nature is in debate. To obtain a full picture in gamma-ray band, we implement a comprehensive targeted search of burst emission in a wide window of 30 days jointly with Insight-HXMT, GECAM and Fermi/GBM data within the ETJASMIN framework. In gamma-ray band, we find there is a 50-second precursor about 25 hours before the 4-hour main burst, which generally consists of 4 emission episodes. Remarkably, we find that the soft X-ray emission (after the main burst) decays as a power-law with start time aligning with the last episode of main emission and index of -5/3 perfectly consistent with the canonical prediction of fallback accretion. We conclude that the properties of precursor, main burst and the following soft X-ray emission strongly support the atypical collapsar Ultra-Long Gamma-Ray Burst (ULGRB) scenario rather than the Tidal Disruption Event (TDE), and all these gamma-ray and soft X-ray emission probably originate from relativistic jet whose luminosity is dominated by the fallback accretion rate during the death collapse of a supergiant star.

astro-ph.HE↗

arXiv2025

LUNCH: A Lightweight Unified Deep-Learning Framework for General Transients Classification in High-Energy Time-Domain Astronomy

The increasing data volume of high-energy space monitors necessitates real-time, automated transient classification for multi-messenger follow-up. Conventional methods rely on empirical features like hardness ratios and reliable localization, which are not always precisely available during early detection. We developed the Lightweight Unified Neural Classifier for High-energy Transients (LUNCH) - an end-to-end deep-learning framework that performs general transient classification directly from raw multi-band light curves, eliminating the need for background subtraction or source localization. Its dual-scale architecture fuses long- and short-scale temporal evolution adaptively. Evaluated on 15 years of Fermi/GBM triggers, the optimal model achieves 97.23% accuracy when trained on complete energy spectra. A lightweight version using only three broad energy bands retains 95.07% accuracy, demonstrating that coarse spectral information fused with temporal context enables robust discrimination. The system significantly outperforms the GBM in-flight classifier on three months of independent test data. Feature visualization reveals well-separated class clusters, confirming physical interpretability. LUNCH combines high accuracy, low computational cost, and instrument-agnostic inputs, offering a practical solution for real-time in-flight processing that enables timely triggers for immediate multi-wavelength and multi-messenger follow-up observations in future time-domain missions.

astro-ph.IM↗

arXiv2025

PolaRiS: Scalable Real-to-Sim Evaluations for Generalist Robot Policies

A significant challenge for robot learning research is our ability to accurately measure and compare the performance of robot policies. Benchmarking in robotics is historically challenging due to the stochasticity, reproducibility, and time-consuming nature of real-world rollouts. This challenge is exacerbated for recent generalist policies, which has to be evaluated across a wide variety of scenes and tasks. Evaluation in simulation offers a scalable complement to real world evaluations, but the visual and physical domain gap between existing simulation benchmarks and the real world has made them an unreliable signal for policy improvement. Furthermore, building realistic and diverse simulated environments has traditionally required significant human effort and expertise. To bridge the gap, we introduce Policy Evaluation and Environment Reconstruction in Simulation (PolaRiS), a scalable real-to-sim framework for high-fidelity simulated robot evaluation. PolaRiS utilizes neural reconstruction methods to turn short video scans of real-world scenes into interactive simulation environments. Additionally, we develop a simple simulation data co-training recipe that bridges remaining real-to-sim gaps and enables zero-shot evaluation in unseen simulation environments. Through extensive paired evaluations between simulation and the real world, we demonstrate that PolaRiS evaluations provide a much stronger correlation to real world generalist policy performance than existing simulated benchmarks. Its simplicity also enables rapid creation of diverse simulated environments. As such, this work takes a step towards distributed and democratized evaluation for the next generation of robotic foundation models.

cs.RO↗

arXiv2025

MobileWorld: Benchmarking Autonomous Mobile Agents in Agent-User Interactive and MCP-Augmented Environments

Among existing online mobile-use benchmarks, AndroidWorld has emerged as the dominant benchmark due to its reproducible environment and deterministic evaluation; however, recent agents achieving over 90% success rates indicate its saturation and motivate the need for a more challenging benchmark. In addition, its environment lacks key application categories, such as e-commerce and enterprise communication, and does not reflect realistic mobile-use scenarios characterized by vague user instructions and hybrid tool usage. We introduce MobileWorld, a substantially more challenging benchmark designed to reflect real-world usage through 201 tasks across 20 applications. MobileWorld derives its difficulty from an emphasis on long-horizon, cross-application workflows, requiring nearly twice as many completion steps on average (27.8 vs. 14.3) and featuring a significantly higher proportion of multi-app tasks (62.2% vs. 9.5%) than AndroidWorld. To overcome the limitations of existing environments, MobileWorld achieves a balance between production-grade utility and reproducible evaluation by utilizing open-source alternatives to industry standards (e.g., Mattermost for Slack). This approach enables a fully observable and controlled environment through source code modification and direct backend database access for precise verification. MobileWorld also introduces novel task categories, including agent-user interaction and Model Context Protocol (MCP)-augmented tasks, for evaluating agents in user-aware, hybrid-tool scenarios. To facilitate evaluation, we develop a planner-executor agentic framework with extended action spaces to support user interactions and MCP calls. Our results reveal a sharp performance drop compared to AndroidWorld, with the best agentic framework and end-to-end model achieving 51.7% and 20.9% success rates, respectively, highlighting ample headroom for future research.

cs.CL↗

arXiv2025

MAI-UI Technical Report: Real-World Centric Foundation GUI Agents

The development of GUI agents could revolutionize the next generation of human-computer interaction. Motivated by this vision, we present MAI-UI, a family of foundation GUI agents spanning the full spectrum of sizes, including 2B, 8B, 32B, and 235B-A22B variants. We identify four key challenges to realistic deployment: the lack of native agent-user interaction, the limits of UI-only operation, the absence of a practical deployment architecture, and brittleness in dynamic environments. MAI-UI addresses these issues with a unified methodology: a self-evolving data pipeline that expands the navigation data to include user interaction and MCP tool calls, a native device-cloud collaboration system routes execution by task state, and an online RL framework with advanced optimizations to scale parallel environments and context length. MAI-UI establishes new state-of-the-art across GUI grounding and mobile navigation. On grounding benchmarks, it reaches 73.5% on ScreenSpot-Pro, 91.3% on MMBench GUI L2, 70.9% on OSWorld-G, and 49.2% on UI-Vision, surpassing Gemini-3-Pro and Seed1.8 on ScreenSpot-Pro. On mobile GUI navigation, it sets a new SOTA of 76.7% on AndroidWorld, surpassing UI-Tars-2, Gemini-2.5-Pro and Seed1.8. On MobileWorld, MAI-UI obtains 41.7% success rate, significantly outperforming end-to-end GUI models and competitive with Gemini-3-Pro based agentic frameworks. Our online RL experiments show significant gains from scaling parallel environments from 32 to 512 (+5.2 points) and increasing environment step budget from 15 to 50 (+4.3 points). Finally, the native device-cloud collaboration system improves on-device performance by 33%, reduces cloud model calls by over 40%, and preserves user privacy.

cs.CV↗

arXiv2025

Physics-informed Diffusion Models for Multi-scale Prediction of Reference Signal Received Power in Wireless Networks

The Reference Signal Received Power (RSRP) is a crucial factor that determines communication performance in mobile networks. Accurately predicting the RSRP can help network operators perceive user experiences and maximize throughput by optimizing wireless resources. However, existing research into RSRP prediction has limitations in accuracy and verisimilitude. Theoretical derivations and existing data-driven methods consider only easily quantifiable Large-Scale (LS) information, and struggle to effectively capture the intertwined LS and Small-Scale (SS) signal attenuation characteristics of the wireless channel. Moreover, the lack of prior physical knowledge leads to weak accuracy, interpretability, and transferability. In this paper, we propose a novel RSRP prediction framework, Channel-Diff. This framework physically models LS and SS attenuation using multimodal conditions and employs physics-informed conditional diffusion models as the prediction network. Channel-Diff extracts prior physical information that characterises the signal propagation process from network parameters and multi-attribute maps of the urban spatial environment. It provides LS physical priors through large-scale propagation modelling and shadow-occlusion modelling, and SS physical priors through multipath propagation modelling and urban microenvironment feature extraction. We design a physical-prior-guided two-stage training scheme with a noise prior guidance mechanism, enabling effective fusion of multi-scale physical knowledge with the diffusion models. Evaluations demonstrate Channel-Diff exhibits excellent performance on RSRP prediction, achieving at least 25.15%-37.19% improvement in accuracy relative to baseline methods. Additionally, the model also demonstrated outstanding performance in terms of transferability and training efficiency.

cs.NI↗

arXiv2025

ETP-R1: Evolving Topological Planning with Reinforcement Fine-tuning for Vision-Language Navigation in Continuous Environments

Vision-Language Navigation in Continuous Environments (VLN-CE) requires an embodied agent to navigate towards target in continuous environments, following natural language instructions. While current graph-based methods offer an efficient, structured approach by abstracting the environment into a topological map and simplifying the action space to waypoint selection, they lag behind methods based on Large Vision-Language Models (LVLMs) in leveraging large-scale data and advanced training paradigms. In this paper, we try to bridge this gap by introducing ETP-R1, a framework that applies the paradigm of scaling up data and Reinforcement Fine-Tuning (RFT) to a graph-based VLN-CE model. To build a strong foundation, we first construct a high-quality, large-scale pretraining dataset using the Gemini API. This dataset consists of diverse, low-hallucination instructions for topological trajectories, providing rich supervision for our graph-based policy to map language to topological paths. This foundation is further strengthened by unifying data from both R2R and RxR tasks for joint pretraining. Building on this, we introduce a three-stage training paradigm, which culminates in the first application of closed-loop, online RFT to a graph-based VLN-CE model, powered by the Group Relative Policy Optimization (GRPO) algorithm. Extensive experiments demonstrate that our approach is highly effective, establishing new state-of-the-art performance across all major metrics on both the R2R-CE and RxR-CE benchmarks. Our code is available at https://github.com/Cepillar/ETP-R1.

cs.RO↗

arXiv2025

Exact boundary controllability and stabilizability of a degenerated Timoshenko beam

This paper investigates the boundary controllability and stabilizability of a Timoshenko beam subject to degeneracy at one end, while control is applied at the opposite boundary. Degeneracy in this context is measured by the real parameters for $μ_a\in [0,2)$ for $a\in\{K,EI\}$, where $K(x)$ denotes shear stiffness and $EI(x)$ bending stiffness. We differentiate between weak degeneracy $μ_a\in [0,1)$ and strong degeneracy $μ_a\in [1,2)$, which may occur independently in shear and bending. Our study establishes observability inequalities for both weakly and strongly degenerate equations under Dirichlet, Robin, and Neumann boundary conditions. Using energy multiplier techniques and the Hilbert Uniqueness Method (HUM), we derive conditions for exact boundary controllability and show that appropriate boundary state and velocity feedback controls at the non-degenerate end can stabilize the system exponentially. Extending results previously obtained for the 1-dimensional wave equation in \cite{AlabauCannarsaLeugering2017}, this study highlights new control strategies and stabilization effects specific to the degenerate Timoshenko beam system, addressing challenges pertinent to real-world structural damping and control applications.

math.OC↗

arXiv2025

Asymmetric exact controllability for networks of spatial elastic strings, springs and masses

We consider networks of elastic strings with end masses, where the coupling is modeled via elastic springs. The model is representative of a network of nonlinear strings, where the strings are coupled to elastic bodies. The coupled system converges to the classical string network model with Kirchhoff and continuity transmission conditions as the spring stiffness terms approach infinity and the masses at the nodes vanish. Due to the presence of point masses at the nodes, the boundary conditions become dynamic, and consequently, the corresponding first-order system of quasilinear balance laws exhibits nonlocal boundary conditions. We demonstrate well-posedness in the sense of semi-global classical solutions \cite{li} (i.e., for arbitrarily large time intervals provided that the initial and boundary data are small enough) and observe extra regularity at the masses as in \cite{WangLeugeringLi2017,WangLeugeringLi2019}. We prove local and global-local exact boundary controllability of a star-like network when control is active at the endpoints of the string-spring-mass system, except for one clamped end. In this case, at multiple nodes, a complex smoothing pattern appears, leading to asymmetric control spaces when the springs and masses are present. Furthermore, the rank of the Laplacian matrix at the junction is crucial for the controllability property, particularly in models containing wave equations with degeneration at dynamic boundaries, which can be interpreted as damage in mechanical vibration systems where parts of the springs are missing.

math.OC↗

arXiv2025

CodeTF: One-stop Transformer Library for State-of-the-art Code LLMs

Code intelligence plays a key role in transforming modern software engineering. Recently, deep learning-based models, especially Transformer-based large language models (LLMs), have demonstrated remarkable potential in tackling these tasks by leveraging massive open-source code data and programming language features. However, the development and deployment of such models often require expertise in both machine learning and software engineering, creating a barrier for the model adoption. In this paper, we present CodeTF, an open-source Transformer-based library for state-of-the-art Code LLMs and code intelligence. Following the principles of modular design and extensible framework, we design CodeTF with a unified interface to enable rapid access and development across different types of models, datasets and tasks. Our library supports a collection of pretrained Code LLM models and popular code benchmarks, including a standardized interface to train and serve code LLMs efficiently, and data features such as language-specific parsers and utility functions for extracting code attributes. In this paper, we describe the design principles, the architecture, key modules and components, and compare with other related library tools. Finally, we hope CodeTF is able to bridge the gap between machine learning/generative AI and software engineering, providing a comprehensive open-source solution for developers, researchers, and practitioners.

cs.SE↗

arXiv2025

Finding the Particularity of the Active Episode of SGR J1935+2154 during Which FRB 20200428 Occurred: Implication from Statistics of Fermi/GBM X-Ray Bursts

By using the Fermi/Gamma-ray Burst Monitor data of the X-ray bursts (XRBs) of SGR J1935+2154, we investigate the temporal clustering of the bursts and the cumulative distribution of the waiting time and fluence/flux. It is found that the bursts occurring in the episode hosting FRB 20200428 have obviously shorter waiting times than those in the other episodes. The general statistical properties of the XRBs further indicate they could belong to a self-organized critical (SOC) system (e.g., starquakes), making them very similar to the earthquake phenomena. Then, according to a unified scaling law between the waiting time and energy of the earthquakes as well as their aftershocks, we implement an analogy analysis on the XRBs and find that the FRB episode owns more dependent burst events than the other episodes. It is indicated that the fast radio burst (FRB) emission could be produced by the interaction between different burst events, which could correspond to a collision between different seismic/Alfven waves or different explosion outflows. Such a situation could appear when the magnetar enters into a global intensive activity period.

astro-ph.HE↗

arXiv2025

GECAM Observations of the Galactic Magnetar SGR J1935+2154 during the 2021 and 2022 Burst Active Episodes. I. Burst Catalog

Magnetar is a neutron star with an ultrahigh magnetic field ($\sim 10^{14}-10^{15}$ G). The magnetar SGR J1935+2154 is not only one of the most active magnetars detected so far, but also the unique confirmed source of fast radio bursts (FRBs). Gravitational wave high-energy Electromagnetic Counterpart All-sky Monitor (GECAM) is dedicated to monitor gamma-ray transients all over the sky, including magnetar short bursts. Here we report the GECAM observations of the burst activity of SGR J1935+2154 from January 2021 to December 2022, which results in a unique and valuable data set for this important magnetar. With a targeted search of GECAM data, 159 bursts from SGR J1935+2154 are detected by GECAM-B while 97 bursts by GECAM-C, including the X-ray burst associated with a bright radio burst. We find that both the burst duration and the waiting time between two successive bursts follow lognormal distributions. The period of burst activity is $134\pm20$ days, thus the burst activity could be generally divided into four active episodes over these two years. Interestingly, the hardness ratio of X-ray bursts tends to be softer during these two years, especially during the active episode with radio bursts detected.

astro-ph.HE↗

arXiv2025

A Random Batch Method for the Efficient Simulation and Optimal Control of Networked 1-D Wave Equations

In this paper, a stochastic algorithm for the efficient simulation and optimal control of networked wave equations based on the random batch method is proposed and analyzed. The random approximation is constructed by dividing the time interval into subintervals and restricting the dynamics to a randomly chosen subnetwork during each of these subintervals. It is proven that the solution for this randomized system converges in expectation to the solution on the original network when the length of the subintervals approaches zero. Furthermore, the optimal controls for the randomized system converge (in $H^2$ and in expectation) to the optimal controls for the original system. The computational advantage of the proposed method is demonstrated in two numerical examples.

math.OC↗

arXiv2025

OmniScientist: Toward a Co-evolving Ecosystem of Human and AI Scientists

With the rapid development of Large Language Models (LLMs), AI agents have demonstrated increasing proficiency in scientific tasks, ranging from hypothesis generation and experimental design to manuscript writing. Such agent systems are commonly referred to as "AI Scientists." However, existing AI Scientists predominantly formulate scientific discovery as a standalone search or optimization problem, overlooking the fact that scientific research is inherently a social and collaborative endeavor. Real-world science relies on a complex scientific infrastructure composed of collaborative mechanisms, contribution attribution, peer review, and structured scientific knowledge networks. Due to the lack of modeling for these critical dimensions, current systems struggle to establish a genuine research ecosystem or interact deeply with the human scientific community. To bridge this gap, we introduce OmniScientist, a framework that explicitly encodes the underlying mechanisms of human research into the AI scientific workflow. OmniScientist not only achieves end-to-end automation across data foundation, literature review, research ideation, experiment automation, scientific writing, and peer review, but also provides comprehensive infrastructural support by simulating the human scientific system, comprising: (1) a structured knowledge system built upon citation networks and conceptual correlations; (2) a collaborative research protocol (OSP), which enables seamless multi-agent collaboration and human researcher participation; and (3) an open evaluation platform (ScienceArena) based on blind pairwise user voting and Elo rankings. This infrastructure empowers agents to not only comprehend and leverage human knowledge systems but also to collaborate and co-evolve, fostering a sustainable and scalable innovation ecosystem.

cs.CY↗

arXiv2025

Towards Efficient and Effective Multi-Camera Encoding for End-to-End Driving

We present Flex, an efficient and effective scene encoder that addresses the computational bottleneck of processing high-volume multi-camera data in end-to-end autonomous driving. Flex employs a small set of learnable scene tokens to jointly encode information from all image tokens across different cameras and timesteps. By design, our approach is geometry-agnostic, learning a compact scene representation directly from data without relying on the explicit 3D inductive biases, such as Bird-Eye-View (BEV), occupancy or tri-plane representations, which are common in prior work. This holistic encoding strategy aggressively compresses the visual input for the downstream Large Language Model (LLM) based policy model. Evaluated on a large-scale proprietary dataset of 20,000 driving hours, our Flex achieves 2.2x greater inference throughput while improving driving performance by a large margin compared to state-of-the-art methods. Furthermore, we show that these compact scene tokens develop an emergent capability for scene decomposition without any explicit supervision. Our findings challenge the prevailing assumption that 3D priors are necessary, demonstrating that a data-driven, joint encoding strategy offers a more scalable, efficient and effective path for future autonomous driving systems.

cs.CV↗

arXiv2025

Seeing to Act, Prompting to Specify: A Bayesian Factorization of Vision Language Action Policy

The pursuit of out-of-distribution generalization in Vision-Language-Action (VLA) models is often hindered by catastrophic forgetting of the Vision-Language Model (VLM) backbone during fine-tuning. While co-training with external reasoning data helps, it requires experienced tuning and data-related overhead. Beyond such external dependencies, we identify an intrinsic cause within VLA datasets: modality imbalance, where language diversity is much lower than visual and action diversity. This imbalance biases the model toward visual shortcuts and language forgetting. To address this, we introduce BayesVLA, a Bayesian factorization that decomposes the policy into a visual-action prior, supporting seeing-to-act, and a language-conditioned likelihood, enabling prompt-to-specify. This inherently preserves generalization and promotes instruction following. We further incorporate pre- and post-contact phases to better leverage pre-trained foundation models. Information-theoretic analysis formally validates our effectiveness in mitigating shortcut learning. Extensive experiments show superior generalization to unseen instructions, objects, and environments compared to existing methods. Project page is available at: https://xukechun.github.io/papers/BayesVLA.

cs.RO↗

arXiv2025

Advancing Identification method of Gamma-Ray Bursts with Data and Feature Enhancement

Gamma-ray bursts (GRBs) are challenging to identify due to their transient nature, complex temporal profiles, and limited observational datasets. We address this with a one-dimensional convolutional neural network integrated with an Adaptive Frequency Feature Enhancement module and physics-informed data augmentation. Our framework generates 100,000 synthetic GRB samples, expanding training data diversity and volume while preserving physical fidelity-especially for low-significance events. The model achieves 97.46% classification accuracy, outperforming all tested variants with conventional enhancement modules, highlighting enhanced domain-specific feature capture. Feature visualization shows model focuses on deep-seated morphological features and confirms the capability of extracting physically meaningful burst characteristics. Dimensionality reduction and clustering reveal GRBs with similar morphologies or progenitor origins cluster in the feature space, linking learned features to physical properties. This perhaps offers a novel diagnostic tool for identifying kilonova- and supernova-associated GRB candidates, establishing criteria to enhance multi-messenger early-warning systems. The framework aids current time-domain surveys, generalizes to other rare transients, and advances automated detection in large-volume observational data.

astro-ph.HE↗