arXiv subjects

Bo Wu

Publications and source records attributed to Bo Wu.

At least 37 records · Page 2Linked to original sources

arXiv2025

Improved Physics-Driven Neural Network to Solve Inverse Scattering Problems

This paper presents an improved physics-driven neural network (IPDNN) framework for solving electromagnetic inverse scattering problems (ISPs). A new Gaussian-localized oscillation-suppressing window (GLOW) activation function is introduced to stabilize convergence and enable a lightweight yet accurate network architecture. A dynamic scatter subregion identification strategy is further developed to adaptively refine the computational domain, preventing missed detections and reducing computational cost. Moreover, transfer learning is incorporated to extend the solver's applicability to practical scenarios, integrating the physical interpretability of iterative algorithms with the real-time inference capability of neural networks. Numerical simulations and experimental results demonstrate that the proposed solver achieves superior reconstruction accuracy, robustness, and efficiency compared with existing state-of-the-art methods.

cs.LG↗

arXiv2025

Near-realtime Earth Observation Via Starlink LEO Satellite Constellation

Earth observation (EO) satellites in Low Earth Orbit (LEO) are collecting vast amounts of data, which are invaluable for applications such as monitoring forest fires. However, data downloading from EO satellites faces significant challenges due to the limited number of ground stations and the brief communication windows with them. Conversely, emerging LEO constellations like Starlink have enabled continuous connectivity and revolutionized access for ordinary users globally, who can connect via a simple satellite dish. In this paper, we study the feasibility of supporting EO satellites with Starlink satellite infrastructure and introduce a novel data delivery system, designated as "Starlink Space User" (SSU), for relaying data from observation satellites. SSU treats EO satellites as space users of Starlink, facilitating efficient data transfer to Earth. At the core of SSU is a novel class of algorithms designed for link and PoP selection, as well as system scheduling optimization, that operate effectively atop Starlink's proprietary infrastructure. We assess the performance of SSU using trace-driven simulations alongside real-world Starlink performance measurements. Our results demonstrate that the proposed Starlink-aided design can significantly reduce the median backlog (data not delivered) per satellite.

cs.NI↗

arXiv2025

VoiceAgentEval: A Dual-Dimensional Benchmark for Expert-Level Intelligent Voice-Agent Evaluation of Xbench's Professional-Aligned Series

We propose OutboundEval, a comprehensive benchmark for evaluating large language models (LLMs) in expert-level intelligent outbound calling scenarios. Unlike existing methods that suffer from three key limitations - insufficient dataset diversity and category coverage, unrealistic user simulation, and inaccurate evaluation metrics - OutboundEval addresses these issues through a structured framework. First, we design a benchmark spanning six major business domains and 30 representative sub-scenarios, each with scenario-specific process decomposition, weighted scoring, and domain-adaptive metrics. Second, we develop a large-model-driven User Simulator that generates diverse, persona-rich virtual users with realistic behaviors, emotional variability, and communication styles, providing a controlled yet authentic testing environment. Third, we introduce a dynamic evaluation method that adapts to task variations, integrating automated and human-in-the-loop assessment to measure task execution accuracy, professional knowledge application, adaptability, and user experience quality. Experiments on 12 state-of-the-art LLMs reveal distinct trade-offs between expert-level task completion and interaction fluency, offering practical insights for building reliable, human-like outbound AI systems. OutboundEval establishes a practical, extensible, and domain-oriented standard for benchmarking LLMs in professional applications.

cs.AI↗

arXiv2025

Investigating Student Interaction Patterns with Large Language Model-Powered Course Assistants in Computer Science Courses

Providing students with flexible and timely academic support is a challenge at most colleges and universities, leaving many students without help outside scheduled hours. Large language models (LLMs) are promising for bridging this gap, but interactions between students and LLMs are rarely overseen by educators. We developed and studied an LLM-powered course assistant deployed across multiple computer science courses to characterize real-world use and understand pedagogical implications. By Spring 2024, our system had been deployed to approximately 2,000 students across six courses at three institutions. Analysis of the interaction data shows that usage remains strong in the evenings and nights and is higher in introductory courses, indicating that our system helps address temporal support gaps and novice learner needs. We sampled 200 conversations per course for manual annotation: most sampled responses were judged correct and helpful, with a small share unhelpful or erroneous; few responses included dedicated examples. We also examined an inquiry-based learning strategy: only around 11% of sampled conversations contained LLM-generated follow-up questions, which were often ignored by students in advanced courses. A Bloom's taxonomy analysis reveals that current LLM capabilities are limited in generating higher-order cognitive questions. These patterns suggest opportunities for pedagogically oriented LLM-based educational systems and greater educator involvement in configuring prompts, content, and policies.

cs.CY↗

arXiv2025

Stage-wise Adaptive Label Distribution for Facial Age Estimation

Label ambiguity poses a significant challenge in age estimation tasks. Most existing methods address this issue by modeling correlations between adjacent age groups through label distribution learning. However, they often overlook the varying degrees of ambiguity present across different age stages. In this paper, we propose a Stage-wise Adaptive Label Distribution Learning (SA-LDL) algorithm, which leverages the observation -- revealed through our analysis of embedding similarities between an anchor and all other ages -- that label ambiguity exhibits clear stage-wise patterns. By jointly employing stage-wise adaptive variance modeling and weighted loss function, SA-LDL effectively captures the complex and structured nature of label ambiguity, leading to more accurate and robust age estimation. Extensive experiments demonstrate that SA-LDL achieves competitive performance, with MAE of 1.74 and 2.15 on the MORPH-II and FG-NET datasets.

cs.CV↗

arXiv2025

Results of the NeurIPS 2023 Neural MMO Competition on Multi-task Reinforcement Learning

We present the results of the NeurIPS 2023 Neural MMO Competition, which attracted over 200 participants and submissions. Participants trained goal-conditional policies that generalize to tasks, maps, and opponents never seen during training. The top solution achieved a score 4x higher than our baseline within 8 hours of training on a single 4090 GPU. We open-source everything relating to Neural MMO and the competition under the MIT license, including the policy weights and training code for our baseline and for the top submissions.

cs.LG↗

arXiv2025

Combinatorial Development of Amorphous/nanocrystalline Biphase Soft Magnetic Alloys with Silicon-steel like Saturated Magnetic Induction

Maximization saturation magnetic induction (Bs) of soft magnetic alloys is essential for the high power-density electromagnetic devices. However, identifying the alloy compositions with high Bs often replies on the lab-intensive melt casting method and a high-throughput characterization on magnetic properties remains challenging. Here, we develop a new combinatorial method for fast screening alloys with optimal soft magnetic properties based on the high-throughput MOKE screening method. Based on the combinatorial method, we found that the alloys with a combination of high Bs and low coercivity (Hc) tend to have a feature of amorphous-nanocrystalline biphase microstructure. We also identified an amorphous/nanocrystalline alloy film with the composition the Fe68.09Co17.02B10.9Si4, exhibiting an ultra-high Bs up to 2.02 T that surpasses all amorphous/nanocrystalline alloys reported so far and is comparable to that of silicon steels, together with a high resistivity of 882 μΩ {\dot} cm, about 17 times of silicon steels. Our high-throughput magnetic screening method provides a paradigm for understanding the relationship between microstructure and magnetic properties and the development of the next-generation soft magnetic materials.

cond-mat.mtrl-sci↗

arXiv2025

Enhancing Interpretability in Software Change Management with Chain-of-Thought Reasoning

In modern online services, frequent software changes introduce significant risks. To tackle this challenge, we propose SCELM (Software Change Evaluation and Lifecycle Management), an end-to-end automated framework for software change management. SCELM aims to manage software changes efficiently and precisely, significantly reducing service failures and economic losses.

cs.SE↗

arXiv2025

LlamaRL: A Distributed Asynchronous Reinforcement Learning Framework for Efficient Large-scale LLM Training

Reinforcement Learning (RL) has become the most effective post-training approach for improving the capabilities of Large Language Models (LLMs). In practice, because of the high demands on latency and memory, it is particularly challenging to develop an efficient RL framework that reliably manages policy models with hundreds to thousands of billions of parameters. In this paper, we present LlamaRL, a fully distributed, asynchronous RL framework optimized for efficient training of large-scale LLMs with various model sizes (8B, 70B, and 405B parameters) on GPU clusters ranging from a handful to thousands of devices. LlamaRL introduces a streamlined, single-controller architecture built entirely on native PyTorch, enabling modularity, ease of use, and seamless scalability to thousands of GPUs. We also provide a theoretical analysis of LlamaRL's efficiency, including a formal proof that its asynchronous design leads to strict RL speed-up. Empirically during the Llama 3 post-training, by leveraging best practices such as colocated model offloading, asynchronous off-policy training, and distributed direct memory access for weight synchronization, LlamaRL achieves significant efficiency gains -- up to 10.7x speed-up compared to DeepSpeed-Chat-like systems on a 405B-parameter policy model. Furthermore, the efficiency advantage continues to grow with increasing model scale, demonstrating the framework's suitability for future large-scale RL training.

cs.LG↗

arXiv2025

ChartGen: Scaling Chart Understanding Via Code-Guided Synthetic Chart Generation

Chart-to-code reconstruction -- the task of recovering executable plotting scripts from chart images -- provides important insights into a model's ability to ground data visualizations in precise, machine-readable form. Yet many existing multimodal benchmarks largely focus primarily on answering questions about charts or summarizing them. To bridge this gap, we present ChartGen, a fully-automated pipeline for code-guided synthetic chart generation. Starting from seed chart images, ChartGen (i) prompts a vision-language model (VLM) to reconstruct each image into a python script, and (ii) iteratively augments that script with a code-oriented large language model (LLM). Using ChartGen, we create 222.5K unique chart-image code pairs from 13K seed chart images, and present an open-source synthetic chart dataset covering 27 chart types, 11 plotting libraries, and multiple data modalities (image, code, text, CSV, DocTags). From this corpus, we curate a held-out chart-to-code evaluation subset of 4.3K chart image-code pairs, and evaluate six open-weight VLMs (3B - 26B parameters), highlighting substantial room for progress. We release the pipeline, prompts, and the dataset to help accelerate efforts towards robust chart understanding and vision-conditioned code generation: https://github.com/SD122025/ChartGen/

cs.HC↗

arXiv2025

R^3-VQA: "Read the Room" by Video Social Reasoning

"Read the room" is a significant social reasoning capability in human daily life. Humans can infer others' mental states from subtle social cues. Previous social reasoning tasks and datasets lack complexity (e.g., simple scenes, basic interactions, incomplete mental state variables, single-step reasoning, etc.) and fall far short of the challenges present in real-life social interactions. In this paper, we contribute a valuable, high-quality, and comprehensive video dataset named R^3-VQA with precise and fine-grained annotations of social events and mental states (i.e., belief, intent, desire, and emotion) as well as corresponding social causal chains in complex social scenarios. Moreover, we include human-annotated and model-generated QAs. Our task R^3-VQA includes three aspects: Social Event Understanding, Mental State Estimation, and Social Causal Reasoning. As a benchmark, we comprehensively evaluate the social reasoning capabilities and consistencies of current state-of-the-art large vision-language models (LVLMs). Comprehensive experiments show that (i) LVLMs are still far from human-level consistent social reasoning in complex social scenarios; (ii) Theory of Mind (ToM) prompting can help LVLMs perform better on social reasoning tasks. We provide some of our dataset and codes in supplementary material and will release our full dataset and codes upon acceptance.

cs.CV↗

arXiv2025

Photonic logic tensor computing beyond TOPS per core

The soaring demand for computing resources has spurred great interest in photonic computing with higher speed and larger computing capacity. Photonic logic gates are of crucial importance due to the fundamental role of Boolean logic in modern digital computing systems. However, most photonic logic schemes struggle to exhibit the capability of massively parallel processing and flexible reconfiguration, owing to weak and fixed nonlinearity in optical elements. Here, we propose a photonic logic tensor computing architecture for the first time and fabricate the photonic universal logic tensor core (PULTC) with a parallel logic computing capacity beyond TOPS. Ten wavelength channels and four spatial channels are designed in PULTC, where the logic computing speed in each channel can reach 50 Gbit/s. After the nonlinear mapping of microring modulators, arbitrary logic operations can be achieved by configuring the Mach-Zehnder interferometer mesh. Our work offers an innovative route for photonic universal logic computing with high-parallel capability and propels the practical applications of photonic logic computing.

physics.optics↗

arXiv2025

WLTCL: Wide Field-of-View 3-D LiDAR Truck Compartment Automatic Localization System

As an essential component of logistics automation, the automated loading system is becoming a critical technology for enhancing operational efficiency and safety. Precise automatic positioning of the truck compartment, which serves as the loading area, is the primary step in automated loading. However, existing methods have difficulty adapting to truck compartments of various sizes, do not establish a unified coordinate system for LiDAR and mobile manipulators, and often exhibit reliability issues in cluttered environments. To address these limitations, our study focuses on achieving precise automatic positioning of key points in large, medium, and small fence-style truck compartments in cluttered scenarios. We propose an innovative wide field-of-view 3-D LiDAR vehicle compartment automatic localization system. For vehicles of various sizes, this system leverages the LiDAR to generate high-density point clouds within an extensive field-of-view range. By incorporating parking area constraints, our vehicle point cloud segmentation method more effectively segments vehicle point clouds within the scene. Our compartment key point positioning algorithm utilizes the geometric features of the compartments to accurately locate the corner points, providing stackable spatial regions. Extensive experiments on our collected data and public datasets demonstrate that this system offers reliable positioning accuracy and reduced computational resource consumption, leading to its application and promotion in relevant fields.

cs.CV↗

arXiv2025

The Influence of Establishing Belt and Road Node Cities on the Development of Digital Inclusive Finance Across Chinese Provinces

Domestic and foreign scholars have conducted many studies on the influencing factors of digital inclusive finance, but few studies on the impact of the "Belt and Road" and the establishment of node cities on digital inclusive finance. Therefore, it is necessary to study the impact of the establishment of node cities in the "Belt and Road" on the development of digital inclusive finance. This paper mainly uses descriptive analysis and literature analysis to summarize and analyze the development status of China's digital inclusive finance and relevant theoretical research on the development of the "Belt and Road" Initiative, and analyzes and collates the background knowledge needed for the demonstration from the aspects of the development status of China's digital inclusive finance and the impact of digital inclusive finance on the economy. In this process, referring to the relevant economic theories, the theoretical model of this paper is proposed and the influence machine analysis is carried out. Empirically, this paper selects the development level of digital inclusive finance in 31 provinces in China from 2011 to 2020 as the explained variable, takes the establishment of "Belt and Road" node cities as a quasi-natural experiment, and verifies the promoting effect of the establishment of "Belt and Road" node cities on the development of digital inclusive finance in provinces through the differential differential method. And verify whether the level of Internet development is a mediating variable. The empirical results show that the establishment of node cities in the "Belt and Road" does promote the development of digital inclusive finance in provinces with the level of Internet development as an intermediary variable.

econ.GN↗

arXiv2025

Short-Term Effects of COVID-19 on Wages: Empirical Evidence and Underlying Mechanisms

This study investigates the causal relationship between the COVID-19 pandemic and wage levels, aiming to provide a quantified assessment of the impact. While no significant evidence is found for long-term effects, the analysis reveals a statistically significant positive influence on wages in the short term, particularly within a one-year horizon. Contrary to common expectations, the results suggest that COVID-19 may have led to short-run wage increases. Several potential mechanisms are proposed to explain this counterintuitive outcome. The findings remain robust when controlling for other macroeconomic indicators such as GDP, considered here as a proxy for aggregate demand. The paper also addresses issues of external validity in the concluding section.

econ.GN↗

arXiv2025

SVLTA: Benchmarking Vision-Language Temporal Alignment via Synthetic Video Situation

Vision-language temporal alignment is a crucial capability for human dynamic recognition and cognition in real-world scenarios. While existing research focuses on capturing vision-language relevance, it faces limitations due to biased temporal distributions, imprecise annotations, and insufficient compositionally. To achieve fair evaluation and comprehensive exploration, our objective is to investigate and evaluate the ability of models to achieve alignment from a temporal perspective, specifically focusing on their capacity to synchronize visual scenarios with linguistic context in a temporally coherent manner. As a preliminary step, we present the statistical analysis of existing benchmarks and reveal the existing challenges from a decomposed perspective. To this end, we introduce SVLTA, the Synthetic Vision-Language Temporal Alignment derived via a well-designed and feasible control generation method within a simulation environment. The approach considers commonsense knowledge, manipulable action, and constrained filtering, which generates reasonable, diverse, and balanced data distributions for diagnostic evaluations. Our experiments reveal diagnostic insights through the evaluations in temporal question answering, distributional shift sensitiveness, and temporal alignment adaptation.

cs.CV↗

arXiv2025

Gemma 3 Technical Report

We introduce Gemma 3, a multimodal addition to the Gemma family of lightweight open models, ranging in scale from 1 to 27 billion parameters. This version introduces vision understanding abilities, a wider coverage of languages and longer context - at least 128K tokens. We also change the architecture of the model to reduce the KV-cache memory that tends to explode with long context. This is achieved by increasing the ratio of local to global attention layers, and keeping the span on local attention short. The Gemma 3 models are trained with distillation and achieve superior performance to Gemma 2 for both pre-trained and instruction finetuned versions. In particular, our novel post-training recipe significantly improves the math, chat, instruction-following and multilingual abilities, making Gemma3-4B-IT competitive with Gemma2-27B-IT and Gemma3-27B-IT comparable to Gemini-1.5-Pro across benchmarks. We release all our models to the community.

cs.CL↗

arXiv2025

Income Inequality, Food Aid, and 'Zero Hunger': Evaluating Effectiveness During Lula's Administration

Income inequality has been an important social issue that has attracted widespread attention. Taking the Zero Hunger Program in Brazil as a case study, this study analyzes the impact of policy changes on the income distribution of the Brazilian population during the implementation of the program using a breakpoint regression approach. The data for the study come from a variety of sources, including the Brazilian Ministry of Development, Social Assistance and the Family, and are analyzed using detailed descriptive statistics from the CEIC Brazil In-Depth Database. The results of the study show that during the Lula administration, Brazil's Zero Hunger Program substantially reduced income inequality, provided more substantial income security for the poor, and reduced the income gap between the rich and the poor. In terms of gender differences, the program led to a larger increase in the income of the male labor force at the right age compared to the female labor force, further highlighting the positive impact of the policy on the male labor force. These results are further confirmed by sensitivity analysis and provide useful lessons for subsequent policy formulation. This study's deep dive into the effects of the Zero Hunger program provides a valuable contribution to academic social science research and policy development.

econ.GN↗