arXiv subjects

Li Zhao

Publications and source records attributed to Li Zhao.

At least 55 records · Page 3Linked to original sources

arXiv2025

HW-GNN: Homophily-Aware Gaussian-Window Constrained Graph Spectral Network for Social Network Bot Detection

Social bots are increasingly polluting online platforms by spreading misinformation and engaging in coordinated manipulation, posing severe threats to cybersecurity. Graph Neural Networks (GNNs) have become mainstream for social bot detection due to their ability to integrate structural and attribute features, with spectral-based approaches demonstrating particular efficacy due to discriminative patterns in the spectral domain. However, current spectral GNN methods face two limitations: (1) their broad-spectrum fitting mechanisms degrade the focus on bot-specific spectral features, and (2) certain domain knowledge valuable for bot detection, e.g., low homophily correlates with high-frequency features, has not been fully incorporated into existing methods. To address these challenges, we propose HW-GNN, a novel homophily-aware graph spectral network with Gaussian window constraints. Our framework introduces two key innovations: (i) a Gaussian-window constrained spectral network that employs learnable Gaussian windows to highlight bot-related spectral features, and (ii) a homophily-aware adaptation mechanism that injects domain knowledge between homophily ratios and frequency features into the Gaussian window optimization process. Through extensive experimentation on multiple benchmark datasets, we demonstrate that HW-GNN achieves state-of-the-art bot detection performance, outperforming existing methods with an average improvement of 4.3% in F1-score, while exhibiting strong plug-in compatibility with existing spectral GNNs.

cs.SI↗

arXiv2025

Discover, Learn, and Reinforce: Scaling Vision-Language-Action Pretraining with Diverse RL-Generated Trajectories

Scaling vision-language-action (VLA) model pre-training requires large volumes of diverse, high-quality manipulation trajectories. Most current data is obtained via human teleoperation, which is expensive and difficult to scale. Reinforcement learning (RL) methods learn useful skills through autonomous exploration, making them a viable approach for generating data. However, standard RL training collapses to a narrow execution pattern, limiting its utility for large-scale pre-training. We propose Discover, Lea rn and Reinforce (DLR), an information-theoretic pattern discovery framework that generates multiple distinct, high-success behavioral patterns for VLA pretraining. Empirically, DLR generates a markedly more diverse trajectory corpus on LIBERO. Specifically, it learns multiple distinct, high-success strategies for the same task where standard RL discovers only one, and hence it covers substantially broader regions of the state-action space. When adapted to unseen downstream task suites, VLA models pretrained on our diverse RL data surpass counterparts trained on equal-sized standard RL datasets. Moreover, DLR exhibits positive data-scaling behavior that single-pattern RL lacks. These results position multi-pattern RL as a practical, scalable data engine for embodied foundation models.

cs.RO↗

arXiv2025

What Do Latent Action Models Actually Learn?

Latent action models (LAMs) aim to learn action-relevant changes from unlabeled videos by compressing changes between frames as latents. However, differences between video frames can be caused by controllable changes as well as exogenous noise, leading to an important concern -- do latents capture the changes caused by actions or irrelevant noise? This paper studies this issue analytically, presenting a linear model that encapsulates the essence of LAM learning, while being tractable.This provides several insights, including connections between LAM and principal component analysis (PCA), desiderata of the data-generating policy, and justification of strategies to encourage learning controllable changes using data augmentation, data cleaning, and auxiliary action-prediction. We also provide illustrative results based on numerical simulation, shedding light on the specific structure of observations, actions, and noise in data that influence LAM learning.

cs.LG↗

arXiv2025

How Do VLAs Effectively Inherit from VLMs?

Vision-language-action (VLA) models hold the promise to attain generalizable embodied control. To achieve this, a pervasive paradigm is to leverage the rich vision-semantic priors of large vision-language models (VLMs). However, the fundamental question persists: How do VLAs effectively inherit the prior knowledge from VLMs? To address this critical question, we introduce a diagnostic benchmark, GrinningFace, an emoji tabletop manipulation task where the robot arm is asked to place objects onto printed emojis corresponding to language instructions. This task design is particularly revealing -- knowledge associated with emojis is ubiquitous in Internet-scale datasets used for VLM pre-training, yet emojis themselves are largely absent from standard robotics datasets. Consequently, they provide a clean proxy: successful task completion indicates effective transfer of VLM priors to embodied control. We implement this diagnostic task in both simulated environment and a real robot, and compare various promising techniques for knowledge transfer. Specifically, we investigate the effects of parameter-efficient fine-tuning, VLM freezing, co-training, predicting discretized actions, and predicting latent actions. Through systematic evaluation, our work not only demonstrates the critical importance of preserving VLM priors for the generalization of VLA but also establishes guidelines for future research in developing truly generalizable embodied AI systems.

cs.RO↗

arXiv2025

Tritiated methane reduction in the PandaX-4T experiment via purge and cryogenic distillation processes

Tritium from tritiated methane (CH$_3$T) calibration is a significant impurity that restricts the sensitivity of the PandaX-4T dark matter detection experiment in the low-energy region. The CH$_3$T removal is essential for PandaX-4T and other liquid xenon dark matter direct detection experiments, as CH$_3$T serves as a critical component for low-energy calibration. To eliminate CH$_3$T, the xenon in the detector is suitably recuperated, leaving 1.8 bar of xenon gas inside, and the detector is flushed with heated xenon gas. Concurrently, leveraging the lower boiling point of methane relative to xenon, the PandaX-4T cryogenic distillation system is effectively utilized to extract CH$_3$T from xenon after optimizing the operational parameters. Following the commissioning run, 5.7 tons of xenon are purified via the distillation method. Recent data indicate that the CH$_3$T concentration reduces from $3.6\times10^{-24}$ mol/mol to $5.9\times10^{-25}$ mol/mol, demonstrating that gas purging and distillation are effective in removing CH$_3$T, even at concentrations on the order of $10^{-24}$ mol/mol.

physics.ins-det↗

arXiv2025

FinResearchBench: A Logic Tree based Agent-as-a-Judge Evaluation Framework for Financial Research Agents

Recently, AI agents are rapidly evolving in intelligence and widely used in professional research applications, such as STEM, software development, and finance. Among these AI agents, deep research agent is a key category as it can perform long-horizon tasks and solve problems of greater complexity. However, there are few evaluation frameworks and benchmarks that systematically and automatically investigate the capabilities of these research agents. In addition, financial research problems have distinct complexity and subtlety. To fill in the gap, we propose FinResearchBench, which is a logic tree-based Agent-as-a-Judge and targets specifically for the financial research agents. It provides a comprehensive and automatic assessment of the research agents across 7 key types of tasks in the financial research domain. The contributions of this work are two-folded: (1) the first and innovative Agent-as-a-Judge system that extracts the logic tree of the research outcome and uses it as the intermediate information to present a comprehensive, reliable, and robust evaluation; (2) finance-oriented that it covers 70 typical financial research questions, spreading across 7 frequently encountered types of task in the domain.

cs.CL↗

arXiv2025

Beyond Human Demonstrations: Diffusion-Based Reinforcement Learning to Generate Data for VLA Training

Vision-language-action (VLA) models have shown strong generalization across tasks and embodiments; however, their reliance on large-scale human demonstrations limits their scalability owing to the cost and effort of manual data collection. Reinforcement learning (RL) offers a potential alternative to generate demonstrations autonomously, yet conventional RL algorithms often struggle on long-horizon manipulation tasks with sparse rewards. In this paper, we propose a modified diffusion policy optimization algorithm to generate high-quality and low-variance trajectories, which contributes to a diffusion RL-powered VLA training pipeline. Our algorithm benefits from not only the high expressiveness of diffusion models to explore complex and diverse behaviors but also the implicit regularization of the iterative denoising process to yield smooth and consistent demonstrations. We evaluate our approach on the LIBERO benchmark, which includes 130 long-horizon manipulation tasks, and show that the generated trajectories are smoother and more consistent than both human demonstrations and those from standard Gaussian RL policies. Further, training a VLA model exclusively on the diffusion RL-generated data achieves an average success rate of 81.9%, which outperforms the model trained on human data by +5.3% and that on Gaussian RL-generated data by +12.6%. The results highlight our diffusion RL as an effective alternative for generating abundant, high-quality, and low-variance demonstrations for VLA models.

cs.RO↗

arXiv2025

Constraints on quantum Oppenheimer-Snyder black holes with eccentric extreme mass-ratio inspirals

We investigate the potential of extreme mass-ratio inspirals to constrain quantum Oppenheimer-Snyder black holes within the framework of loop quantum gravity. We consider a stellar-mass object orbiting a supermassive Oppenheimer-Snyder black hole in an equatorial eccentric trajectory. To explore the dynamical behavior of the system, we analyze its orbital evolution under gravitational radiation within the adiabatic approximation and the mass-quadrupole formula for different initial orbital configurations. Our results show that the quantum correction parameter $\hatα$ slows down the evolution of the orbital semi-latus rectum and eccentricity. We then employ the numerical kludge method to generate the corresponding time-domain gravitational waveforms. To assess detectability, we include Doppler modulation due to the motion of space-based detectors and compute the frequency-domain characteristic strain. By evaluating mismatches between response signals for different values of $\hatα$, we show that even small corrections $(\hatα \sim 10^{-5})$ produce distinguishable effects. Our analysis suggests that future space-based detectors such as LISA can probe quantum gravitational corrections in the strong-field regime and place constraints significantly stronger than those from black hole shadow observations.

gr-qc↗

arXiv2025

Infrastructure Sensor-enabled Vehicle Data Generation using Multi-Sensor Fusion for Proactive Safety Applications at Work Zone

Infrastructure-based sensing and real-time trajectory generation show promise for improving safety in high-risk roadway segments such as work zones, yet practical deployments are hindered by perspective distortion, complex geometry, occlusions, and costs. This study tackles these barriers by integrating roadside camera and LiDAR sensors into a cosimulation environment to develop a scalable, cost-effective vehicle detection and localization framework, and employing a Kalman Filter-based late fusion strategy to enhance trajectory consistency and accuracy. In simulation, the fusion algorithm reduced longitudinal error by up to 70 percent compared to individual sensors while preserving lateral accuracy within 1 to 3 meters. Field validation in an active work zone, using LiDAR, a radar-camera rig, and RTK-GPS as ground truth, demonstrated that the fused trajectories closely match real vehicle paths, even when single-sensor data are intermittent or degraded. These results confirm that KF based sensor fusion can reliably compensate for individual sensor limitations, providing precise and robust vehicle tracking capabilities. Our approach thus offers a practical pathway to deploy infrastructure-enabled multi-sensor systems for proactive safety measures in complex traffic environments.

cs.CV↗

arXiv2025

villa-X: Enhancing Latent Action Modeling in Vision-Language-Action Models

Vision-Language-Action (VLA) models have emerged as a popular paradigm for learning robot manipulation policies that can follow language instructions and generalize to novel scenarios. Recent works have begun to explore the incorporation of latent actions, abstract representations of motion between two frames, into VLA pre-training. In this paper, we introduce villa-X, a novel Vision-Language-Latent-Action (ViLLA) framework that advances latent action modeling for learning generalizable robot manipulation policies. Our approach improves both how latent actions are learned and how they are incorporated into VLA pre-training. We demonstrate that villa-X can generate latent action plans in a zero-shot fashion, even for unseen embodiments and open-vocabulary symbolic understanding. This capability enables villa-X to achieve superior performance across diverse simulation tasks in SIMPLER and on two real-world robotic setups involving both gripper and dexterous hand manipulation. These results establish villa-X as a principled and scalable paradigm for learning generalizable robot manipulation policies. We believe it provides a strong foundation for future research.

cs.RO↗

arXiv2025

HyKid: An Open MRI Dataset with Expert-Annotated Multi-Structure and Choroid Plexus in Pediatric Hydrocephalus

Evaluation of hydrocephalus in children is challenging, and the related research is limited by a lack of publicly available, expert-annotated datasets, particularly those with segmentation of the choroid plexus. To address this, we present HyKid, an open-source dataset from 48 pediatric patients with hydrocephalus. 3D MRIs were provided with 1mm isotropic resolution, which was reconstructed from routine low-resolution images using a slice-to-volume algorithm. Manually corrected segmentations of brain tissues, including white matter, grey matter, lateral ventricle, external CSF, and the choroid plexus, were provided by an experienced neurologist. Additionally, structured data was extracted from clinical radiology reports using a Retrieval-Augmented Generation framework. The strong correlation between choroid plexus volume and total CSF volume provided a potential biomarker for hydrocephalus evaluation, achieving excellent performance in a predictive model (AUC = 0.87). The proposed HyKid dataset provided a high-quality benchmark for neuroimaging algorithms development, and it revealed the choroid plexus-related features in hydrocephalus assessments. Our datasets are publicly available at https://www.synapse.org/Synapse:syn68544889.

cs.CV↗

arXiv2025

SAM-TTT: Segment Anything Model via Reverse Parameter Configuration and Test-Time Training for Camouflaged Object Detection

This paper introduces a new Segment Anything Model (SAM) that leverages reverse parameter configuration and test-time training to enhance its performance on Camouflaged Object Detection (COD), named SAM-TTT. While most existing SAM-based COD models primarily focus on enhancing SAM by extracting favorable features and amplifying its advantageous parameters, a crucial gap is identified: insufficient attention to adverse parameters that impair SAM's semantic understanding in downstream tasks. To tackle this issue, the Reverse SAM Parameter Configuration Module is proposed to effectively mitigate the influence of adverse parameters in a train-free manner by configuring SAM's parameters. Building on this foundation, the T-Visioner Module is unveiled to strengthen advantageous parameters by integrating Test-Time Training layers, originally developed for language tasks, into vision tasks. Test-Time Training layers represent a new class of sequence modeling layers characterized by linear complexity and an expressive hidden state. By integrating two modules, SAM-TTT simultaneously suppresses adverse parameters while reinforcing advantageous ones, significantly improving SAM's semantic understanding in COD task. Our experimental results on various COD benchmarks demonstrate that the proposed approach achieves state-of-the-art performance, setting a new benchmark in the field. The code will be available at https://github.com/guobaoxiao/SAM-TTT.

cs.CV↗

arXiv2025

Infrastructure-enabled risk assessment of hazardous road conditions on rural roads during inclement weather

Rural roadways often expose Commercial Motor Vehicle (CMV) drivers to hazardous conditions, such as heavy fog, rain, snow, black ice, and flash floods, many of which remain unreported in real time. This lack of timely information, coupled with limited infrastructure in rural areas, significantly increases the risk of crashes. Although various sensing technologies exist to monitor individual hazards like low visibility or surface friction, they rarely assess the combined driving risk posed by multiple simultaneous hazards, nor do they provide actionable recommendations such as safe advisory speeds. To address this critical gap, in this study, we present a roadway hazard risk assessment framework that provides an approach to quantify the probability and severity of crash occurrences due to specific roadway hazards. To evaluate this framework, we presented a case study by constructing a synthetic "year-long" dataset that encompasses every possible pairing of road surface and visibility conditions. Our analysis confirms that the combined ProbabilitySeverity scoring yields a coherent, stepwise risk profile across all hazard scenarios. These results validate the practicality of our risk assessment approach and provide a foundation for deploying graduated safety measures in real-world roadway operations.

cs.CE↗

arXiv2025

Historical Prediction Attention Mechanism based Trajectory Forecasting for Proactive Work Zone Safety in a Digital Twin Environment

Proactive safety systems aim to mitigate risks by anticipating potential conflicts between vehicles and enabling early intervention to prevent work zone-related crashes. This study presents an infrastructure-enabled proactive work zone safety warning system that leverages a Digital Twin environment, integrating real-time multi-sensor data, detailed High-Definition (HD) maps, and a historical prediction attention mechanism-based trajectory prediction model. Using a co-simulation environment that combines Simulation of Urban MObility (SUMO) and CAR Learning to Act (CARLA) simulators, along with Lanelet2 HD maps and the Historical Prediction Network (HPNet) model, we demonstrate effective trajectory prediction and early warning generation for vehicle interactions in freeway work zones. To evaluate the accuracy of predicted trajectories, we use two standard metrics: Joint Average Displacement Error (ADE) and Joint Final Displacement Error (FDE). Specifically, the infrastructure-enabled HPNet model demonstrates superior performance on the work-zone datasets generated from the co-simulation environment, achieving a minimum Joint FDE of 0.3228 meters and a minimum Joint ADE of 0.1327 meters, lower than the benchmarks on the Argoverse (minJointFDE: 1.0986 m, minJointADE: 0.7612 m) and Interaction (minJointFDE: 0.8231 m, minJointADE: 0.2548 m) datasets. In addition, our proactive safety warning generation application, utilizing vehicle bounding boxes and probabilistic conflict modeling, demonstrates its capability to issue alerts for potential vehicle conflicts.

cs.OH↗

arXiv2025

Lessons Learned from the Real-World Deployment of Multi-Sensor Fusion for Proactive Work Zone Safety Application

Proactive safety systems that anticipate and mitigate traffic risks before incidents occur are increasingly recognized as essential for improving work zone safety. Unlike traditional reactive methods, these systems rely on real-time sensing, trajectory prediction, and intelligent infrastructure to detect potential hazards. Existing simulation-based studies often overlook, and real-world deployment studies rarely discuss the practical challenges associated with deploying such systems in operational settings, particularly those involving roadside infrastructure and multi-sensor integration and fusion. This study addresses that gap by presenting deployment insights and technical lessons learned from a real-world implementation of a multi-sensor safety system at an active bridge repair work zone along the N-2/US-75 corridor in Lincoln, Nebraska. The deployed system combines LiDAR, radar, and camera sensors with an edge computing platform to support multi-modal object tracking, trajectory fusion, and real-time analytics. Specifically, this study presents key lessons learned across three critical stages of deployment: (1) sensor selection and placement, (2) sensor calibration, system integration, and validation, and (3) sensor fusion. Additionally, we propose a predictive digital twin framework that leverages fused trajectory data for early conflict detection and real-time warning generation, enabling proactive safety interventions.

cs.OH↗

arXiv2025

PIG-Nav: Key Insights for Pretrained Image Goal Navigation Models

Recent studies have explored pretrained (foundation) models for vision-based robotic navigation, aiming to achieve generalizable navigation and positive transfer across diverse environments while enhancing zero-shot performance in unseen settings. In this work, we introduce PIG-Nav (Pretrained Image-Goal Navigation), a new approach that further investigates pretraining strategies for vision-based navigation models and contributes in two key areas. Model-wise, we identify two critical design choices that consistently improve the performance of pretrained navigation models: (1) integrating an early-fusion network structure to combine visual observations and goal images via appropriately pretrained Vision Transformer (ViT) image encoder, and (2) introducing suitable auxiliary tasks to enhance global navigation representation learning, thus further improving navigation performance. Dataset-wise, we propose a novel data preprocessing pipeline for efficiently labeling large-scale game video datasets for navigation model training. We demonstrate that augmenting existing open navigation datasets with diverse gameplay videos improves model performance. Our model achieves an average improvement of 22.6% in zero-shot settings and a 37.5% improvement in fine-tuning settings over existing visual navigation foundation models in two complex simulated environments and one real-world environment. These results advance the state-of-the-art in pretrained image-goal navigation models. Notably, our model maintains competitive performance while requiring significantly less fine-tuning data, highlighting its potential for real-world deployment with minimal labeled supervision.

cs.CV↗

arXiv2025

Dyn-O: Building Structured World Models with Object-Centric Representations

World models aim to capture the dynamics of the environment, enabling agents to predict and plan for future states. In most scenarios of interest, the dynamics are highly centered on interactions among objects within the environment. This motivates the development of world models that operate on object-centric rather than monolithic representations, with the goal of more effectively capturing environment dynamics and enhancing compositional generalization. However, the development of object-centric world models has largely been explored in environments with limited visual complexity (such as basic geometries). It remains underexplored whether such models can generalize to more complex settings with diverse textures and cluttered scenes. In this paper, we fill this gap by introducing Dyn-O, an enhanced structured world model built upon object-centric representations. Compared to prior work in object-centric representations, Dyn-O improves in both learning representations and modeling dynamics. On the challenging Procgen games, we find that our method can learn object-centric world models directly from pixel observations, outperforming DreamerV3 in rollout prediction accuracy. Furthermore, by decoupling object-centric features into dynamics-agnostic and dynamics-aware components, we enable finer-grained manipulation of these features and generate more diverse imagined trajectories.

cs.LG↗

arXiv2025

Policy Filtration for RLHF to Mitigate Noise in Reward Models

While direct policy optimization methods exist, pioneering LLMs are fine-tuned with reinforcement learning from human feedback (RLHF) to generate better responses under the supervision of a reward model learned from preference data. One major challenge of RLHF is the inaccuracy of the intermediate reward model, especially in the tasks that requires complex reasoning for the reward model to score a response. We find that the reliability of the reward model varies across responses assigned with different rewards. This motivates us to filter the samples whose rewards may be unreliable to improve the signal-to-noise ratio during policy learning, resulting in Policy Filtration for Proximal Policy Optimization (PF-PPO). To choose a proper policy filtering strategy, we use the coefficient of determination (R2) between the rewards and actual scores on filtered samples as the metrics to help us find promising strategies since it measures how well the rewards filtered by PF-PPO indicate real performance. We provide extensive experiments to validate the effectiveness of PF-PPO in code generation and math reasoning tasks. In code generation, PF-PPO achieves the state-of-the-art performance of 7-billion-parameter models on HumanEval (+7.9%), MBPP (+0.7%), and LeetCode Contest (+10.0%) which is a more challenging benchmark created by us. In math reasoning, PF-PPO yields performance increase using different reward models and benchmarks (Ape210K and CMATH). Code is available on https://github.com/DtYXs/verl/tree/pf-ppo.

cs.LG↗