arXiv subjects

Chen Wang

Publications and source records attributed to Chen Wang.

At least 253 records · Page 14Linked to original sources

arXiv2025

The Fire We Share

The Fire We Share proposes a care-centered, consequence-aware visualization framework for engaging with wildfire data not as static metrics, but as living archives of ecological and social entanglement. By combining plants-inspired data forms, event-based mapping, and narrative layering, the project foregrounds fire as a shared temporal condition-one that cuts across natural cycles and human systems. Rather than simplifying wildfire data into digestible visuals, The Fire We Share reimagines it as a textured, wounded archive-embodied, relational, and radically ethical.

cs.GR↗

arXiv2025

EFRame: Deeper Reasoning via Exploration-Filter-Replay Reinforcement Learning Framework

Recent advances in reinforcement learning (RL) have significantly enhanced the reasoning capabilities of large language models (LLMs). Group Relative Policy Optimization (GRPO), a lightweight variant of Proximal Policy Optimization (PPO), improves efficiency but suffers from limited exploration and training instability, limiting its effectiveness on complex reasoning tasks. To address these challenges, we introduce EFRame, an Exploration-Filter-Replay framework that augments GRPO across three dimensions: additional rollouts enable deeper and more targeted exploration, online filtering removes low-quality samples to stabilize gradients and accelerate training, and experience replay amplifies rare yet informative trajectories for stable convergence. This unified framework establishes a principled training cycle that balances exploration, efficiency, and stability. Experiments on diverse reasoning benchmarks demonstrate that EFRame achieves consistent gains, including a 37.9\% relative improvement on Geometry3K over GRPO. EFRame further supports fine-grained sample categorization and precise entropy control, highlighting it as a robust solution for advancing deeper reasoning in LLMs. Our code is available at https://github.com/597358816/EFRame.

cs.LG↗

arXiv2025

Ergodicity and error estimate of laws for a random splitting Langevin Monte Carlo

The random splitting Langevin Monte Carlo could mitigate the first order bias in Langevin Monte Carlo with little extra work compared other high order schemes. We develop in this work an analysis framework for the sampling error under Wasserstein distance regarding the random splitting Langevin Monte Carlo. First, the sharp local truncation error is obtained by the relative entropy approach together with the explicit formulas for the commutator of related semi-groups. The necessary pointwise estimates of the gradient and Hessian of the logarithmic density are established by the Bernstein type approach in PDE theory. Second, the geometric ergodicity is established by accommodation of the reflection coupling. Combining the ergodicity with the local error estimate, we establish a uniform-in-time sampling error bound, showing that the invariant measure of the method approximates the true Gibbs distribution with $O(τ^2)$ accuracy where $τ$ is the time step. Lastly, we perform numerical experiments to validate the theoretical results.

math.NA↗

arXiv2025

When to Reason: Semantic Router for vLLM

Large Language Models (LLMs) demonstrate substantial accuracy gains when augmented with reasoning modes such as chain-of-thought and inference-time scaling. However, reasoning also incurs significant costs in inference latency and token usage, with environmental and financial impacts, which are unnecessary for many simple prompts. We present a semantic router that classifies queries based on their reasoning requirements and selectively applies reasoning only when beneficial. Our approach achieves a 10.2 percentage point improvement in accuracy on the MMLU-Pro benchmark while reducing response latency by 47.1% and token consumption by 48.5% compared to direct inference with vLLM. These results demonstrate that semantic routing offers an effective mechanism for striking a balance between accuracy and efficiency in open-source LLM serving systems

cs.ET↗

arXiv2025

Ambiguity in LLMs is a concept missing problem

Ambiguity in natural language is a significant obstacle for achieving accurate text to structured data mapping through large language models (LLMs), which affects the performance of tasks such as mapping text to agentic tool calling and text-to-SQL queries. Existing methods to ambiguity handling either rely on the ReACT framework to obtain correct mappings through trial and error, or on supervised fine-tuning to bias models toward specific tasks. In this paper, we adopt a different approach that characterizes representation differences of ambiguous text in the latent space and leverages these differences to identify ambiguity before mapping them to structured data. To detect sentence-level ambiguity, we focus on the relationship between ambiguous questions and their interpretations. Unlike distances calculated by dense embeddings, we introduce a new distance measure based on a path kernel over concepts. With this measurement, we identify patterns to distinguish ambiguous from unambiguous questions. Furthermore, we propose a method for improving LLM performance on ambiguous agentic tool calling through missing concept prediction. Both achieve state-of-the-art results.

cs.CL↗

arXiv2025

Data Fusion for Full-Range Response Reconstruction via Diffusion Models

Accurately capturing the full-range response of structures is crucial in structural health monitoring (SHM) for ensuring safety and operational integrity. However, limited sensor deployment due to cost, accessibility, or scale often hinders comprehensive monitoring. This paper presents a generative data fusion framework utilizing diffusion models, to reconstruct the full-range structural response from sparse and heterogeneous sensor measurements. We incorporate Diffusion Posterior Sampling (DPS) into the reconstruction framework, using sensor measurements as probabilistic constraints to guide the sampling process. Three forward models are designed: Direct Observation Mapping (DOM), Channel-based Observation Mapping (COM), and Neural Network Forward Model (NNFM), enabling flexible adaptation to different sensor placement conditions and reconstruction targets. The proposed framework is validated on a steel plate shear wall exhibiting nonlinear responses. By simultaneously sampling 100 realizations and averaging them as the ensemble prediction result, the three forward models achieve Weighted Mean Absolute Percentage Errors of 1.62% (DOM), 3.27% (COM), and 3.49% (NNFM). Sensitivity analyses further demonstrate robust performance under varying hyperparameters, sensor configurations, and noise levels. The proposed framework shows new possibilities for probabilistic modeling and decision-making in SHM by harnessing the capabilities of diffusion models, offering a novel data fusion approach for full-range monitoring of structures.

cs.CE↗

arXiv2025

VStyle: A Benchmark for Voice Style Adaptation with Spoken Instructions

Spoken language models (SLMs) have emerged as a unified paradigm for speech understanding and generation, enabling natural human machine interaction. However, while most progress has focused on semantic accuracy and instruction following, the ability of SLMs to adapt their speaking style based on spoken instructions has received limited attention. We introduce Voice Style Adaptation (VSA), a new task that examines whether SLMs can modify their speaking style, such as timbre, prosody, or persona following natural language spoken commands. To study this task, we present VStyle, a bilingual (Chinese & English) benchmark covering four categories of speech generation: acoustic attributes, natural language instruction, role play, and implicit empathy. We also introduce the Large Audio Language Model as a Judge (LALM as a Judge) framework, which progressively evaluates outputs along textual faithfulness, style adherence, and naturalness, ensuring reproducible and objective assessment. Experiments on commercial systems and open source SLMs demonstrate that current models face clear limitations in controllable style adaptation, highlighting both the novelty and challenge of this task. By releasing VStyle and its evaluation toolkit, we aim to provide the community with a foundation for advancing human centered spoken interaction. The dataset and code are publicly available at \href{https://junzhan2000.github.io/VStyle.github.io/}{project's homepage}.

cs.SD↗

arXiv2025

Intention-aware Hierarchical Diffusion Model for Long-term Trajectory Anomaly Detection

Long-term trajectory anomaly detection is a challenging problem due to the diversity and complex spatiotemporal dependencies in trajectory data. Existing trajectory anomaly detection methods fail to simultaneously consider both the high-level intentions of agents as well as the low-level details of the agent's navigation when analysing an agent's trajectories. This limits their ability to capture the full diversity of normal trajectories. In this paper, we propose an unsupervised trajectory anomaly detection method named Intention-aware Hierarchical Diffusion model (IHiD), which detects anomalies through both high-level intent evaluation and low-level sub-trajectory analysis. Our approach leverages Inverse Q Learning as the high-level model to assess whether a selected subgoal aligns with an agent's intention based on predicted Q-values. Meanwhile, a diffusion model serves as the low-level model to generate sub-trajectories conditioned on subgoal information, with anomaly detection based on reconstruction error. By integrating both models, IHiD effectively utilises subgoal transition knowledge and is designed to capture the diverse distribution of normal trajectories. Our experiments show that the proposed method IHiD achieves up to 30.2% improvement in anomaly detection performance in terms of F1 score over state-of-the-art baselines.

cs.AI↗

arXiv2025

Look Again, Think Slowly: Enhancing Visual Reflection in Vision-Language Models

Recent advances in text-only "slow-thinking" reasoning have prompted efforts to transfer this capability to vision-language models (VLMs), for training visual reasoning models (\textbf{VRMs}). owever, such transfer faces critical challenges: Effective "slow thinking" in VRMs requires \textbf{visual reflection}, the ability to check the reasoning process based on visual information. Through quantitative analysis, we observe that current VRMs exhibit limited visual reflection, as their attention to visual information diminishes rapidly with longer generated responses. To address this challenge, we propose a new VRM \textbf{Reflection-V}, which enhances visual reflection based on reasoning data construction for cold-start and reward design for reinforcement learning (RL). Firstly, we construct vision-centered reasoning data by leveraging an agent that interacts between VLMs and reasoning LLMs, enabling cold-start learning of visual reflection patterns. Secondly, a visual attention based reward model is employed during RL to encourage reasoning based on visual information. Therefore, \textbf{Reflection-V} demonstrates significant improvements across multiple visual reasoning benchmarks. Furthermore, \textbf{Reflection-V} maintains a stronger and more consistent reliance on visual information during visual reasoning, indicating effective enhancement in visual reflection capabilities.

cs.CV↗

arXiv2025

Demystifying Strange O VI Line Widths with Hydrodynamic Simulations

We present a survey of O VI line widths obtained from hydrodynamic simulations that model the mixing between High Velocity Clouds and the circumgalactic medium and track the non-equilibrium ionization populations of the ions. We run 10 simulations with various physical conditions of the clouds and ambient environments, so that our results can be compared to observations of various cloud environments. Synthetic spectra are created for the simulated sight lines that contain O VI. The range of our Doppler broadening parameter, $b$, is from $\sim$6 km s$^{-1}$ to $\sim$107 km s$^{-1}$. We calculate the thermal and non-thermal contributions to the $b$ values. Both $b$ and the thermal contribution to $b$ can be substantially less than the CIE value. These narrow line widths are due to the time delay in recombination from O VI to O V in mixing regions, which results in substantial amounts of O VI at temperatures below the 2.9 $\times$ 10$^{5}$ K CIE temperature. Our results show that non equilibrium collisional ionization and mixing can produce the narrow line widths that are seen in multiple observations.

astro-ph.GA↗

arXiv2025

BagIt! An Adaptive Dual-Arm Manipulation of Fabric Bags for Object Bagging

Bagging tasks, commonly found in industrial scenarios, are challenging considering deformable bags' complicated and unpredictable nature. This paper presents an automated bagging system from the proposed adaptive Structure-of-Interest (SOI) manipulation strategy for dual robot arms. The system dynamically adjusts its actions based on real-time visual feedback, removing the need for pre-existing knowledge of bag properties. Our framework incorporates Gaussian Mixture Models (GMM) for estimating SOI states, optimization techniques for SOI generation, motion planning via Constrained Bidirectional Rapidly-exploring Random Tree (CBiRRT), and dual-arm coordination using Model Predictive Control (MPC). Extensive experiments validate the capability of our system to perform precise and robust bagging across various objects, showcasing its adaptability. This work offers a new solution for robotic deformable object manipulation (DOM), particularly in automated bagging tasks. Video of this work is available at https://youtu.be/6JWjCOeTGiQ.

cs.RO↗

arXiv2025

Spatially-Enhanced Recurrent Memory for Long-Range Mapless Navigation via End-to-End Reinforcement Learning

Recent advancements in robot navigation, particularly with end-to-end learning approaches such as reinforcement learning (RL), have demonstrated strong performance. However, successful navigation still depends on two key capabilities: mapping and planning (explicitly or implicitly). Classical approaches rely on explicit mapping pipelines to register egocentric observations into a coherent map. In contrast, end-to-end learning often achieves this implicitly -- through recurrent neural networks (RNNs) that fuse current and historical observations into a latent space for planning. While existing architectures, such as LSTM and GRU, can capture temporal dependencies, our findings reveal a critical limitation: their inability to effectively perform spatial memorization. This capability is essential for integrating sequential observations from varying perspectives to build spatial representations that support planning. To address this, we propose Spatially-Enhanced Recurrent Units (SRUs) -- a simple yet effective modification to existing RNNs -- that enhance spatial memorization. We further introduce an attention-based network architecture integrated with SRUs, enabling long-range mapless navigation using a single forward-facing stereo camera. We also employ regularization techniques to facilitate robust end-to-end recurrent training via RL. Experimental results show 23.5% overall improvement in long-range navigation compared to existing RNNs. With SRU memory, our method outperforms RL baselines -- one relying on explicit mapping and the other on stacked historical observations -- by 29.6% and 105.0%, respectively, across diverse environments requiring long-horizon mapping and memorization. Finally, we address the sim-to-real gap by leveraging large-scale pretraining on synthetic depth data, enabling zero-shot transfer for deployment across diverse and complex real-world environments.

cs.RO↗

arXiv2025

Nonclassical correlations and quadrature squeezing of photons in anisotropic quantum Rabi-Stark model

This study investigates nonclassical effects of photons in the anisotropic quantum Rabi-Stark model by using a quantum dressed master equation. We analyze second- and higher-order correlation functions, and demonstrate that the nonlinear Stark coupling significantly modulates photon statistics, inducing distinct and tunable photon antibunching and bunching effects. Successive transition signatures of correlation functions provide a potential experimental probe for predicting quantum phase transitions. We further reveal that Stark coupling provides direct control over photon squeezing, achieving significant enhancement and suppression. These findings not only uncover rich nonclassical phenomena in the anisotropic quantum Rabi-Stark model but also establish the nonlinear Stark coupling as a crucial new dimension for quantum detection. It may open a new avenue for precise manipulation of strongly coupled light-matter systems, with potential applications in quantum information processing and quantum-enhanced technologies.

quant-ph↗

arXiv2025

Pixie: Fast and Generalizable Supervised Learning of 3D Physics from Pixels

Inferring the physical properties of 3D scenes from visual information is a critical yet challenging task for creating interactive and realistic virtual worlds. While humans intuitively grasp material characteristics such as elasticity or stiffness, existing methods often rely on slow, per-scene optimization, limiting their generalizability and application. To address this problem, we introduce PIXIE, a novel method that trains a generalizable neural network to predict physical properties across multiple scenes from 3D visual features purely using supervised losses. Once trained, our feed-forward network can perform fast inference of plausible material fields, which coupled with a learned static scene representation like Gaussian Splatting enables realistic physics simulation under external forces. To facilitate this research, we also collected PIXIEVERSE, one of the largest known datasets of paired 3D assets and physic material annotations. Extensive evaluations demonstrate that PIXIE is about 1.46-4.39x better and orders of magnitude faster than test-time optimization methods. By leveraging pretrained visual features like CLIP, our method can also zero-shot generalize to real-world scenes despite only ever been trained on synthetic data. https://pixie-3d.github.io/

cs.CV↗

arXiv2025

Text to Query Plans for Question Answering on Large Tables

Efficient querying and analysis of large tabular datasets remain significant challenges, especially for users without expertise in programming languages like SQL. Text-to-SQL approaches have shown promising performance on benchmark data; however, they inherit SQL's drawbacks, including inefficiency with large datasets and limited support for complex data analyses beyond basic querying. We propose a novel framework that transforms natural language queries into query plans. Our solution is implemented outside traditional databases, allowing us to support classical SQL commands while avoiding SQL's inherent limitations. Additionally, we enable complex analytical functions, such as principal component analysis and anomaly detection, providing greater flexibility and extensibility than traditional SQL capabilities. We leverage LLMs to iteratively interpret queries and construct operation sequences, addressing computational complexity by incrementally building solutions. By executing operations directly on the data, we overcome context length limitations without requiring the entire dataset to be processed by the model. We validate our framework through experiments on both standard databases and large scientific tables, demonstrating its effectiveness in handling extensive datasets and performing sophisticated data analyses.

cs.DB↗

arXiv2025

BEHAVIOR Robot Suite: Streamlining Real-World Whole-Body Manipulation for Everyday Household Activities

Real-world household tasks present significant challenges for mobile manipulation robots. An analysis of existing robotics benchmarks reveals that successful task performance hinges on three key whole-body control capabilities: bimanual coordination, stable and precise navigation, and extensive end-effector reachability. Achieving these capabilities requires careful hardware design, but the resulting system complexity further complicates visuomotor policy learning. To address these challenges, we introduce the BEHAVIOR Robot Suite (BRS), a comprehensive framework for whole-body manipulation in diverse household tasks. Built on a bimanual, wheeled robot with a 4-DoF torso, BRS integrates a cost-effective whole-body teleoperation interface for data collection and a novel algorithm for learning whole-body visuomotor policies. We evaluate BRS on five challenging household tasks that not only emphasize the three core capabilities but also introduce additional complexities, such as long-range navigation, interaction with articulated and deformable objects, and manipulation in confined spaces. We believe that BRS's integrated robotic embodiment, data collection interface, and learning framework mark a significant step toward enabling real-world whole-body manipulation for everyday household tasks. BRS is open-sourced at https://behavior-robot-suite.github.io/

cs.RO↗

arXiv2025

How Much Does Home Field Advantage Matter in Soccer Games? A Causal Inference Approach for English Premier League Analysis

In many sports, it is commonly believed that the home team has an advantage over the visiting team, known as the home field advantage. Yet its causal effect on team performance is largely unknown. In this paper, we propose a novel causal inference approach to study the causal effect of home field advantage in English Premier League. We develop a hierarchical causal model and show that both league level and team level causal effects are identifiable and can be conveniently estimated. We further develop an inference procedure for the proposed estimators and demonstrate its excellent numerical performance via simulation studies. We implement our method on the 2020-21 English Premier League data and assess the causal effect of home advantage on eleven summary statistics that measure the offensive and defensive performance and referee-assessed disciplinary outcomes. We find that the home field advantage resides more heavily in offensive statistics than it does in defensive or referee statistics. We also find evidence that teams that had lower rankings retain a higher home field advantage.

stat.AP↗

arXiv2025

The contribution of electron and hole conductivity to the transport loss in organic solar cells

The effective conductivity determines the reciprocal of the transport resistance, the dominant loss of fill factor in organic solar cells. We experimentally determine the dependence of effective conductivity on its electron and hole contributions. Using PM6:Y12 blends with tunable morphological and energetic disorder, we show that the effective conductivity follows a harmonic mean of electron and hole conductivities even across nearly three orders of magnitude in conductivity imbalance. We also validate the method for directly extracting effective conductivity from current-voltage measurements, eliminating the need to rely on indirect mobility and charge carrier density-based proxies. Our findings challenge the widespread use of geometric mean approximations and offer a more accurate framework for analysing and modelling transport in disordered organic semiconductors.

cond-mat.mtrl-sci↗