arXiv subjects

Bin Xu

Publications and source records attributed to Bin Xu.

At least 73 records · Page 4Linked to original sources

arXiv2025

The Hurwitz existence problem and the prime-degree conjecture: A computational perspective

We investigate the Hurwitz existence problem from a computational viewpoint. Leveraging the symmetric-group algorithm by Zheng and building upon implementations originally developed by Baroni, we achieve a complete and non-redundant enumeration of all non-realizable partition triples for positive integers up to $31$. These results are further categorized into four types according to their underlying mathematical structure; it is observed that nearly nine-tenths of them can be explained by known theoretical results. As an application, we verify the prime-degree conjecture for all primes less than $32$. In light of the exponential memory growth inherent in existing computational approaches -- which limits their feasibility at higher degrees -- we propose a novel software architecture designed to stabilize memory usage, thereby facilitating further detection of exceptional cases in the Hurwitz existence problem. The complete dataset of non-realizable partition triples, along with our implementation, will been made public on GitHub.

math.GR↗

arXiv2025

Hardware Software Optimizations for Fast Model Recovery on Reconfigurable Architectures

Model Recovery (MR) is a core primitive for physical AI and real-time digital twins, but GPUs often execute MR inefficiently due to iterative dependencies, kernel-launch overheads, underutilized memory bandwidth, and high data-movement latency. We present MERINDA, an FPGA-accelerated MR framework that restructures computation as a streaming dataflow pipeline. MERINDA exploits on-chip locality through BRAM tiling, fixed-point kernels, and the concurrent use of LUT fabric and carry-chain adders to expose fine-grained spatial parallelism while minimizing off-chip traffic. This hardware-aware formulation removes synchronization bottlenecks and sustains high throughput across the iterative updates in MR. On representative MR workloads, MERINDA delivers up to 6.3x fewer cycles than an FPGA-based LTC baseline, enabling real-time performance for time-critical physical systems.

cs.AR↗

arXiv2025

Evaluating Hydro-Science and Engineering Knowledge of Large Language Models

Hydro-Science and Engineering (Hydro-SE) is a critical and irreplaceable domain that secures human water supply, generates clean hydropower energy, and mitigates flood and drought disasters. Featuring multiple engineering objectives, Hydro-SE is an inherently interdisciplinary domain that integrates scientific knowledge with engineering expertise. This integration necessitates extensive expert collaboration in decision-making, which poses difficulties for intelligence. With the rapid advancement of large language models (LLMs), their potential application in the Hydro-SE domain is being increasingly explored. However, the knowledge and application abilities of LLMs in Hydro-SE have not been sufficiently evaluated. To address this issue, we propose the Hydro-SE LLM evaluation benchmark (Hydro-SE Bench), which contains 4,000 multiple-choice questions. Hydro-SE Bench covers nine subfields and enables evaluation of LLMs in aspects of basic conceptual knowledge, engineering application ability, and reasoning and calculation ability. The evaluation results on Hydro-SE Bench show that the accuracy values vary among 0.74 to 0.80 for commercial LLMs, and among 0.41 to 0.68 for small-parameter LLMs. While LLMs perform well in subfields closely related to natural and physical sciences, they struggle with domain-specific knowledge such as industry standards and hydraulic structures. Model scaling mainly improves reasoning and calculation abilities, but there is still great potential for LLMs to better handle problems in practical engineering application. This study highlights the strengths and weaknesses of LLMs for Hydro-SE tasks, providing model developers with clear training targets and Hydro-SE researchers with practical guidance for applying LLMs.

cs.CL↗

arXiv2025

Model Recovery at the Edge under Resource Constraints for Physical AI

Model Recovery (MR) enables safe, explainable decision making in mission-critical autonomous systems (MCAS) by learning governing dynamical equations, but its deployment on edge devices is hindered by the iterative nature of neural ordinary differential equations (NODEs), which are inefficient on FPGAs. Memory and energy consumption are the main concerns when applying MR on edge devices for real-time operation. We propose MERINDA, a novel FPGA-accelerated MR framework that replaces iterative solvers with a parallelizable neural architecture equivalent to NODEs. MERINDA achieves nearly 11x lower DRAM usage and 2.2x faster runtime compared to mobile GPUs. Experiments reveal an inverse relationship between memory and energy at fixed accuracy, highlighting MERINDA's suitability for resource-constrained, real-time MCAS.

cs.AI↗

arXiv2025

WebVIA: A Web-based Vision-Language Agentic Framework for Interactive and Verifiable UI-to-Code Generation

User interface (UI) development requires translating design mockups into functional code, a process that remains repetitive and labor-intensive. While recent Vision-Language Models (VLMs) automate UI-to-Code generation, they generate only static HTML/CSS/JavaScript layouts lacking interactivity. To address this, we propose WebVIA, the first agentic framework for interactive UI-to-Code generation and validation. The framework comprises three components: 1) an exploration agent to capture multi-state UI screenshots; 2) a UI2Code model that generates executable interactive code; 3) a validation module that verifies the interactivity. Experiments demonstrate that WebVIA-Agent achieves more stable and accurate UI exploration than general-purpose agents (e.g., Gemini-2.5-Pro). In addition, our fine-tuned WebVIA-UI2Code models exhibit substantial improvements in generating executable and interactive HTML/CSS/JavaScript code, outperforming their base counterparts across both interactive and static UI2Code benchmarks. Our code and models are available at \href{https://zheny2751-dotcom.github.io/webvia.github.io/}{\texttt{https://webvia.github.io}}.

cs.SE↗

arXiv2025

Non-equilibrium Magnon Engineering Enabling Significant Thermal Transport Modulation

Thermal conductivity, a fundamental parameter characterizing thermal transport in solids, is typically determined by electron and phonon transport. Although other transport properties including electrical conductivity and thermoelectric conversion coefficients have material-specific values, it is known that thermal conductivity can be modulated artificially via phonon engineering techniques. Here, we demonstrate another way of artificially modulating the heat conduction in solids: magnonic thermal transport engineering. The time-domain thermoreflectance measurements using ferromagnetic metal/insulator junction systems reveal that the thermal conductivity of the ferromagnetic metals and interfacial thermal conductance vary significantly depending on the spatial distribution of nonequilibrium spin currents. Systematic measurements of the thermal transport properties with changing the boundary conditions for spin currents show that the observed thermal transport modulation stems from magnon origin. This observation unveils that magnons significantly contribute to the heat conduction even in ferromagnetic metals at room temperature, upsetting the conventional wisdom that the thermal conductivity mediated by magnons is very small in metals except at low temperatures. The magnonic thermal transport engineering offers a new principle and method for active thermal management.

cond-mat.mtrl-sci↗

arXiv2025

WebSeer: Training Deeper Search Agents through Reinforcement Learning with Self-Reflection

Search agents have achieved significant advancements in enabling intelligent information retrieval and decision-making within interactive environments. Although reinforcement learning has been employed to train agentic models capable of more dynamic interactive retrieval, existing methods are limited by shallow tool-use depth and the accumulation of errors over multiple iterative interactions. In this paper, we present WebSeer, a more intelligent search agent trained via reinforcement learning enhanced with a self-reflection mechanism. Specifically, we construct a large dataset annotated with reflection patterns and design a two-stage training framework that unifies cold start and reinforcement learning within the self-reflection paradigm for real-world web-based environments, which enables the model to generate longer and more reflective tool-use trajectories. Our approach substantially extends tool-use chains and improves answer accuracy. Using a single 14B model, we achieve state-of-the-art results on HotpotQA and SimpleQA, with accuracies of 72.3% and 90.0%, respectively, and demonstrate strong generalization to out-of-distribution datasets. The code is available at https://github.com/99hgz/WebSeer

cs.CL↗

arXiv2025

High-resolution Photo Enhancement in Real-time: A Laplacian Pyramid Network

Photo enhancement plays a crucial role in augmenting the visual aesthetics of a photograph. In recent years, photo enhancement methods have either focused on enhancement performance, producing powerful models that cannot be deployed on edge devices, or prioritized computational efficiency, resulting in inadequate performance for real-world applications. To this end, this paper introduces a pyramid network called LLF-LUT++, which integrates global and local operators through closed-form Laplacian pyramid decomposition and reconstruction. This approach enables fast processing of high-resolution images while also achieving excellent performance. Specifically, we utilize an image-adaptive 3D LUT that capitalizes on the global tonal characteristics of downsampled images, while incorporating two distinct weight fusion strategies to achieve coarse global image enhancement. To implement this strategy, we designed a spatial-frequency transformer weight predictor that effectively extracts the desired distinct weights by leveraging frequency features. Additionally, we apply local Laplacian filters to adaptively refine edge details in high-frequency components. After meticulously redesigning the network structure and transformer model, LLF-LUT++ not only achieves a 2.64 dB improvement in PSNR on the HDR+ dataset, but also further reduces runtime, with 4K resolution images processed in just 13 ms on a single GPU. Extensive experimental results on two benchmark datasets further show that the proposed approach performs favorably compared to state-of-the-art methods. The source code will be made publicly available at https://github.com/fengzhang427/LLF-LUT.

cs.CV↗

arXiv2025

LoSiA: Efficient High-Rank Fine-Tuning via Subnet Localization and Optimization

Parameter-Efficient Fine-Tuning (PEFT) methods, such as LoRA, significantly reduce the number of trainable parameters by introducing low-rank decomposition matrices. However, existing methods perform extensive matrix multiplications in domain specialization tasks, resulting in computational inefficiency and sub-optimal fine-tuning performance. Hence, we propose LoSiA(Low-Resources Subnet Integration Adaptation), an innovative method that dynamically localizes and optimizes critical parameters during the training process. Specifically, it identifies a sub-network using gradient sparsity analysis and optimizes it as the trainable target. This design enables effective high-rank adaptation by updating only the sub-network parameters, reducing the additional matrix multiplication. We also present LoSiA-Pro, a faster implementation of LoSiA, which reduces the training latency by about $27\%$ compared to LoRA. Extensive evaluations show that our method achieves minimal performance drop compared to full fine-tuning, while requiring the least training time across domain specialization and common-sense reasoning tasks. Further analysis shows that LoSiA also reduces forgetting during continued training. The source code is available at https://github.com/KlozeWang/LoSiA.

cs.LG↗

arXiv2025

Learning in Context: Personalizing Educational Content with Large Language Models to Enhance Student Learning

Standardized, one-size-fits-all educational content often fails to connect with students' individual backgrounds and interests, leading to disengagement and a perceived lack of relevance. To address this challenge, we introduce PAGE, a novel framework that leverages large language models (LLMs) to automatically personalize educational materials by adapting them to each student's unique context, such as their major and personal interests. To validate our approach, we deployed PAGE in a semester-long intelligent tutoring system and conducted a user study to evaluate its impact in an authentic educational setting. Our findings show that students who received personalized content demonstrated significantly improved learning outcomes and reported higher levels of engagement, perceived relevance, and trust compared to those who used standardized materials. This work demonstrates the practical value of LLM-powered personalization and offers key design implications for creating more effective, engaging, and trustworthy educational experiences.

cs.HC↗

arXiv2025

Solution to SU(n+1) Toda system generated by spherical metrics

Using the correspondence between solutions to the SU(n+1) Toda system on a Riemann surface and totally unramified unitary curves, we show that a spherical metric $ω$ generates a family of solutions, including $(i(n+1-i)ω)_{i=1}^n$. Moreover, we characterize this family in terms of the monodromy group of the spherical metric. As a consequence, we obtain a new solution class to the SU(n+1) Toda system with cone singularities on compact Riemann surfaces, complementing the existence results of Lin-Yang-Zhong (JDG, 114(2):337-391, 2020).

math-ph↗

arXiv2025

Leveraging OS-Level Primitives for Robotic Action Management

End-to-end imitation learning frameworks (e.g., VLA) are increasingly prominent in robotics, as they enable rapid task transfer by learning directly from perception to control, eliminating the need for complex hand-crafted features. However, even when employing SOTA VLA-based models, they still exhibit limited generalization capabilities and suboptimal action efficiency, due to the constraints imposed by insufficient robotic training datasets. In addition to addressing this problem using model-based approaches, we observe that robotic action slices, which consist of contiguous action steps, exhibit strong analogies to the time slices of threads in traditional operating systems. This insight presents a novel opportunity to tackle the problem at the system level. In this paper, we propose AMS, a robot action management system enhanced with OS-level primitives like exception, context switch and record-and-replay, that improves both execution efficiency and success rates of robotic tasks. AMS first introduces action exception, which facilitates the immediate interruption of robotic actions to prevent error propagation. Secondly, AMS proposes action context, which eliminates redundant computations for VLA-based models, thereby accelerating execution efficiency in robotic actions. Finally, AMS leverages action replay to facilitate repetitive or similar robotic tasks without the need for re-training efforts. We implement AMS in both an emulated environment and on a real robot platform. The evaluation results demonstrate that AMS significantly enhances the model's generalization ability and action efficiency, achieving task success rate improvements ranging from 7x to 24x and saving end-to-end execution time ranging from 29% to 74% compared to existing robotic system without AMS support.

cs.OS↗

arXiv2025

LVBench: An Extreme Long Video Understanding Benchmark

Recent progress in multimodal large language models has markedly enhanced the understanding of short videos (typically under one minute), and several evaluation datasets have emerged accordingly. However, these advancements fall short of meeting the demands of real-world applications such as embodied intelligence for long-term decision-making, in-depth movie reviews and discussions, and live sports commentary, all of which require comprehension of long videos spanning several hours. To address this gap, we introduce LVBench, a benchmark specifically designed for long video understanding. Our dataset comprises publicly sourced videos and encompasses a diverse set of tasks aimed at long video comprehension and information extraction. LVBench is designed to challenge multimodal models to demonstrate long-term memory and extended comprehension capabilities. Our extensive evaluations reveal that current multimodal models still underperform on these demanding long video understanding tasks. Through LVBench, we aim to spur the development of more advanced models capable of tackling the complexities of long video comprehension. Our data and code are publicly available at: https://lvbench.github.io.

cs.CV↗

arXiv2025

Channel-Independent Federated Traffic Prediction

In recent years, traffic prediction has achieved remarkable success and has become an integral component of intelligent transportation systems. However, traffic data is typically distributed among multiple data owners, and privacy constraints prevent the direct utilization of these isolated datasets for traffic prediction. Most existing federated traffic prediction methods focus on designing communication mechanisms that allow models to leverage information from other clients in order to improve prediction accuracy. Unfortunately, such approaches often incur substantial communication overhead, and the resulting transmission delays significantly slow down the training process. As the volume of traffic data continues to grow, this issue becomes increasingly critical, making the resource consumption of current methods unsustainable. To address this challenge, we propose a novel variable relationship modeling paradigm for federated traffic prediction, termed the Channel-Independent Paradigm(CIP). Unlike traditional approaches, CIP eliminates the need for inter-client communication by enabling each node to perform efficient and accurate predictions using only local information. Based on the CIP, we further develop Fed-CI, an efficient federated learning framework, allowing each client to process its own data independently while effectively mitigating the information loss caused by the lack of direct data sharing among clients. Fed-CI significantly reduces communication overhead, accelerates the training process, and achieves state-of-the-art performance while complying with privacy regulations. Extensive experiments on multiple real-world datasets demonstrate that Fed-CI consistently outperforms existing methods across all datasets and federated settings. It achieves improvements of 8%, 14%, and 16% in RMSE, MAE, and MAPE, respectively, while also substantially reducing communication costs.

cs.LG↗

arXiv2025

Gauss-Bonnet formula for metrics with logarithmic singularities on compact Riemann surfaces

We prove a generalization of the classical Gauss-Bonnet formula for metrics with logarithmic singularities on compact Riemann surfaces, under the condition that the Gaussian curvature is Lebesgue integrable with respect to the metric's area form. As an application, we establish the Gauss-Bonnet formula for special Kähler metrics when their associated cubic differentials are meromorphic on compact Riemann surfaces.

math.DG↗

arXiv2025

SDHN: Skewness-Driven Hypergraph Networks for Enhanced Localized Multi-Robot Coordination

Multi-Agent Reinforcement Learning is widely used for multi-robot coordination, where simple graphs typically model pairwise interactions. However, such representations fail to capture higher-order collaborations, limiting effectiveness in complex tasks. While hypergraph-based approaches enhance cooperation, existing methods often generate arbitrary hypergraph structures and lack adaptability to environmental uncertainties. To address these challenges, we propose the Skewness-Driven Hypergraph Network (SDHN), which employs stochastic Bernoulli hyperedges to explicitly model higher-order multi-robot interactions. By introducing a skewness loss, SDHN promotes an efficient structure with Small-Hyperedge Dominant Hypergraph, allowing robots to prioritize localized synchronization while still adhering to the overall information, similar to human coordination. Extensive experiments on Moving Agents in Formation and Robotic Warehouse tasks validate SDHN's effectiveness, demonstrating superior performance over state-of-the-art baselines.

cs.RO↗

arXiv2025

Identifying the Quadrupolar Nature of Gravitational Wave Background through Space-based Missions

The stochastic gravitational wave background (SGWB) consists of an incoherent collection of waves from both astrophysical and cosmological sources. To distinguish the SGWB from noise, it is essential to verify its quadrupolar nature, exemplified by the cross-correlations among pairs of pulsars within a pulsar timing array, commonly referred to as the Hellings-Downs curve. We study how this quadrupolar signature manifests in correlations between general GW detector pairs, characterized by their antenna responses. Focusing on space-based missions including laser interferometers (LISA, Taiji, TianQin) and atom interferometers (AEDGE/MAGIS), we demonstrate how orbital motion dynamically modulates detector correlations through time-dependent separations and relative orientation shifts. These modulations encode unique statistical features that serve as definitive markers of the SGWB's quadrupolar nature. Our findings identify optimal configurations for these missions, offer forecasts for the time needed to identify the quadrupolar nature of the SGWB, and are applicable to both space-space and space-terrestrial correlations.

gr-qc↗

arXiv2025

Counting Rational Curves and Standard Complex Structures on HyperKähler ALE 4-manifolds

All hyperKähler ALE 4-manifolds with a given non-trivial finite group $Γ$ in $SU(2)$ at infinity are parameterized by an open dense subset of a real linear space of dimension $3$rank$Φ$. Here, $Φ$ denotes the root system associated with $Γ$ via the McKay correspondence. Such manifolds are diffeomorphic to the minimal resolution of a Kleinian singularity. By using the period map of the twistor space, we specify those points in the parameter space at which the hyperKählerian family of complex structures includes the complex structure of the minimal resolution. Furthermore, we count the rational curves lying on each hyperKähler ALE 4-manifold. For each point in the parameter space, we can assign an integer equals to the number of complex structures which contains rational curves. We show this integer function on the parameter space is lower semi-continuous. In the end, based on known results, we prove that the twistor space of any hyperKähler ALE cannot be Kählerian. In particular, we strengthen some results of Kronheimer (J. Differential Geom., 29(3):665--683, 1989) and provide examples of non-compact and non-Kählerian twistor spaces.

math.DG↗