Searcharxiv⌕ Search

arXiv subjects

Xuan Zhao

Publications and source records attributed to Xuan Zhao.

At least 55 records · Page 3Linked to original sources
arXiv2025

X-IL: Exploring the Design Space of Imitation Learning Policies

Designing modern imitation learning (IL) policies requires making numerous decisions, including the selection of feature encoding, architecture, policy representation, and more. As the field rapidly advances, the range of available options continues to grow, creating a vast and largely unexplored design space for IL policies. In this work, we present X-IL, an accessible open-source framework designed to systematically explore this design space. The framework's modular design enables seamless swapping of policy components, such as backbones (e.g., Transformer, Mamba, xLSTM) and policy optimization techniques (e.g., Score-matching, Flow-matching). This flexibility facilitates comprehensive experimentation and has led to the discovery of novel policy configurations that outperform existing methods on recent robot learning benchmarks. Our experiments demonstrate not only significant performance gains but also provide valuable insights into the strengths and weaknesses of various design choices. This study serves as both a practical reference for practitioners and a foundation for guiding future research in imitation learning.

cs.RO↗
arXiv2025

Differentiable Projection-based Learn to Optimize in Wireless Network-Part I: Convex Constrained (Non-)Convex Programming

This paper addresses a class of (non-)convex optimization problems subject to general convex constraints, which pose significant challenges for traditional methods due to their inherent non-convexity and diversity. Conventional convex optimization-based solvers often struggle to efficiently handle these problems in their most general form. While neural network (NN)-based approaches offer a promising alternative, ensuring the feasibility of NN-generated solutions and effectively training the NN remain key hurdles, largely because finite-capacity networks can produce infeasible outputs. To overcome these issues, we propose a projection-based method that projects any infeasible NN output onto the feasible domain, thus guaranteeing strict adherence to the constraints without compromising the NN's optimization capability. Furthermore, we derive the objective function values for both the raw NN outputs and their projected counterparts, along with the gradients of these values with respect to the NN parameters. This derivation enables label-free (unsupervised) training, reducing reliance on labeled data and improving scalability. Experimental results demonstrate that the proposed projection-based method consistently ensures feasibility.

eess.SY↗
arXiv2024

HIGGS: HIerarchy-Guided Graph Stream Summarization

Graph stream summarization refers to the process of processing a continuous stream of edges that form a rapidly evolving graph. The primary challenges in handling graph streams include the impracticality of fully storing the ever-growing datasets and the complexity of supporting graph queries that involve both topological and temporal information. Recent advancements, such as PGSS and Horae, address these limitations by using domain-based, top-down multi-layer structures in the form of compressed matrices. However, they either suffer from poor query accuracy, incur substantial space overheads, or have low query efficiency. This study proposes a novel item-based, bottom-up hierarchical structure, called HIGGS. Unlike existing approaches, HIGGS leverages its hierarchical structure to localize storage and query processing, thereby confining changes and hash conflicts to small and manageable subtrees, yielding notable performance improvements. HIGGS offers tighter theoretical bounds on query accuracy and space cost. Extensive empirical studies on real graph streams demonstrate that, compared to state-of-the-art methods, HIGGS is capable of notable performance enhancements: it can improve accuracy by over 3 orders of magnitude, reduce space overhead by an average of 30%, increase throughput by more than 5 times, and decrease query latency by nearly 2 orders of magnitude.

cs.DB↗
arXiv2024

A unified framework on the original energy laws of three effective classes of Runge-Kutta methods for phase field crystal type models

The main theoretical obstacle to establish the original energy dissipation laws of Runge-Kutta methods for phase-field equations is to verify the maximum norm boundedness of the stage solutions without assuming global Lipschitz continuity of the nonlinear bulk. We present a unified theoretical framework for the energy stability of three effective classes of Runge-Kutta methods, including the additive implicit-explicit Runge-Kutta, explicit exponential Runge-Kutta and corrected integrating factor Runge-Kutta methods, for the Swift-Hohenberg and phase field crystal models. By the standard discrete energy argument, it is proven that the three classes of Runge-Kutta methods preserve the original energy dissipation laws if the associated differentiation matrices are positive definite. Our main tools include the differential form with the associated differentiation matrix, the discrete orthogonal convolution kernels and the principle of mathematical induction. Many existing Runge-Kutta methods in the literature are revisited by evaluating the lower bound on the minimum eigenvalues of the associated differentiation matrices. Our theoretical approach paves a new way for the internal nonlinear stability of Runge-Kutta methods for dissipative semilinear parabolic problems.

math.NA↗
arXiv2024

"I Like Sunnie More Than I Expected!": Exploring User Expectation and Perception of an Anthropomorphic LLM-based Conversational Agent for Well-Being Support

The human-computer interaction (HCI) research community has a longstanding interest in exploring the mismatch between users' actual experiences and expectation toward new technologies, for instance, large language models (LLMs). In this study, we compared users' (N = 38) initial expectations against their post-interaction perceptions of two LLM-powered mental well-being intervention activity recommendation systems. Both systems have a built-in LLM to recommend a personalized well-being intervention activity, but one system (Sunnie) has an anthropomorphic conversational interaction design via elements such as appearance, persona, and natural conversation. Results showed that user engagement was high with both systems, and both systems exceeded users' expectations along the utility dimension, highlighting AI's potential to offer useful intervention activity recommendations. In addition, Sunnie further outperformed the non-anthropomorphic baseline system in relational warmth. These findings suggest that anthropomorphic conversational interaction design may be particularly effective in fostering warmth in mental health support contexts.

cs.HC↗
arXiv2024

Enhancing Fairness through Reweighting: A Path to Attain the Sufficiency Rule

We introduce an innovative approach to enhancing the empirical risk minimization (ERM) process in model training through a refined reweighting scheme of the training data to enhance fairness. This scheme aims to uphold the sufficiency rule in fairness by ensuring that optimal predictors maintain consistency across diverse sub-groups. We employ a bilevel formulation to address this challenge, wherein we explore sample reweighting strategies. Unlike conventional methods that hinge on model size, our formulation bases generalization complexity on the space of sample weights. We discretize the weights to improve training speed. Empirical validation of our method showcases its effectiveness and robustness, revealing a consistent improvement in the balance between prediction performance and fairness metrics across various experiments.

cs.LG↗
arXiv2024

Interlayer Fermi polarons of excited exciton states in quantizing magnetic fields

The study of exciton-polarons has offered profound insights into the many-body interactions between bosonic excitations and their immersed Fermi sea within layered heterostructures. However, little is known about the properties of exciton polarons with interlayer interactions. Here through magneto-optical reflectance contrast measurements, we experimentally investigate interlayer Fermi polarons for 2s excitons in WSe$_2$/graphene heterostructures, where the excited exciton states (2s) in the WSe$_2$ layer are dressed by free charge carriers of the adjacent graphene layer in the Landau quantization regime. First, such a system enables an optical detection of integer and fractional quantum Hall states (e.g. $ν=\pm1/3$, $\pm$2/3) of monolayer graphene. Furthermore, we observe that the 2s state evolves into two distinct branches, denoted as attractive and repulsive polarons, when graphene is doped out of the incompressible quantum Hall gaps. Our work paves the way for the understanding of the excited composite quasiparticles and Bose-Fermi mixtures.

cond-mat.mes-hall↗
arXiv2023

Discovering Galaxy Features via Dataset Distillation

In many applications, Neural Nets (NNs) have classification performance on par or even exceeding human capacity. Moreover, it is likely that NNs leverage underlying features that might differ from those humans perceive to classify. Can we "reverse-engineer" pertinent features to enhance our scientific understanding? Here, we apply this idea to the notoriously difficult task of galaxy classification: NNs have reached high performance for this task, but what does a neural net (NN) "see" when it classifies galaxies? Are there morphological features that the human eye might overlook that could help with the task and provide new insights? Can we visualize tracers of early evolution, or additionally incorporated spectral data? We present a novel way to summarize and visualize galaxy morphology through the lens of neural networks, leveraging Dataset Distillation, a recent deep-learning methodology with the primary objective to distill knowledge from a large dataset and condense it into a compact synthetic dataset, such that a model trained on this synthetic dataset achieves performance comparable to a model trained on the full dataset. We curate a class-balanced, medium-size high-confidence version of the Galaxy Zoo 2 dataset, and proceed with dataset distillation from our accurate NN-classifier to create synthesized prototypical images of galaxy morphological features, demonstrating its effectiveness. Of independent interest, we introduce a self-adaptive version of the state-of-the-art Matching Trajectory algorithm to automate the distillation process, and show enhanced performance on computer vision benchmarks.

cs.CV↗
arXiv2023

Counterfactual Explanation for Regression via Disentanglement in Latent Space

Counterfactual Explanations (CEs) help address the question: How can the factors that influence the prediction of a predictive model be changed to achieve a more favorable outcome from a user's perspective? Thus, they bear the potential to guide the user's interaction with AI systems since they represent easy-to-understand explanations. To be applicable, CEs need to be realistic and actionable. In the literature, various methods have been proposed to generate CEs. However, the majority of research on CEs focuses on classification problems where questions like "What should I do to get my rejected loan approved?" are raised. In practice, answering questions like "What should I do to increase my salary?" are of a more regressive nature. In this paper, we introduce a novel method to generate CEs for a pre-trained regressor by first disentangling the label-relevant from the label-irrelevant dimensions in the latent space. CEs are then generated by combining the label-irrelevant dimensions and the predefined output. The intuition behind this approach is that the ideal counterfactual search should focus on the label-irrelevant characteristics of the input and suggest changes toward target-relevant characteristics. Searching in the latent space could help achieve this goal. We show that our method maintains the characteristics of the query sample during the counterfactual search. In various experiments, we demonstrate that the proposed method is competitive based on different quality measures on image and tabular datasets in regression problem settings. It efficiently returns results closer to the original data manifold compared to three state-of-the-art methods, which is essential for realistic high-dimensional machine learning applications. Our code will be made available as an open-source package upon the publication of this work.

cs.LG↗
arXiv2023

Counterfactual Explanation via Search in Gaussian Mixture Distributed Latent Space

Counterfactual Explanations (CEs) are an important tool in Algorithmic Recourse for addressing two questions: 1. What are the crucial factors that led to an automated prediction/decision? 2. How can these factors be changed to achieve a more favorable outcome from a user's perspective? Thus, guiding the user's interaction with AI systems by proposing easy-to-understand explanations and easy-to-attain feasible changes is essential for the trustworthy adoption and long-term acceptance of AI systems. In the literature, various methods have been proposed to generate CEs, and different quality measures have been suggested to evaluate these methods. However, the generation of CEs is usually computationally expensive, and the resulting suggestions are unrealistic and thus non-actionable. In this paper, we introduce a new method to generate CEs for a pre-trained binary classifier by first shaping the latent space of an autoencoder to be a mixture of Gaussian distributions. CEs are then generated in latent space by linear interpolation between the query sample and the centroid of the target class. We show that our method maintains the characteristics of the input sample during the counterfactual search. In various experiments, we show that the proposed method is competitive based on different quality measures on image and tabular datasets -- efficiently returns results that are closer to the original data manifold compared to three state-of-the-art methods, which are essential for realistic high-dimensional machine learning applications.

cs.LG↗
arXiv2023

Adversarial Reweighting Guided by Wasserstein Distance for Bias Mitigation

The unequal representation of different groups in a sample population can lead to discrimination of minority groups when machine learning models make automated decisions. To address these issues, fairness-aware machine learning jointly optimizes two (or more) metrics aiming at predictive effectiveness and low unfairness. However, the inherent under-representation of minorities in the data makes the disparate treatment of subpopulations less noticeable and difficult to deal with during learning. In this paper, we propose a novel adversarial reweighting method to address such \emph{representation bias}. To balance the data distribution between the majority and the minority groups, our approach deemphasizes samples from the majority group. To minimize empirical risk, our method prefers samples from the majority group that are close to the minority group as evaluated by the Wasserstein distance. Our theoretical analysis shows the effectiveness of our adversarial reweighting approach. Experiments demonstrate that our approach mitigates bias without sacrificing classification accuracy, outperforming related state-of-the-art methods on image and tabular benchmark datasets.

cs.LG↗
arXiv2023

Causal Fairness-Guided Dataset Reweighting using Neural Networks

The importance of achieving fairness in machine learning models cannot be overstated. Recent research has pointed out that fairness should be examined from a causal perspective, and several fairness notions based on the on Pearl's causal framework have been proposed. In this paper, we construct a reweighting scheme of datasets to address causal fairness. Our approach aims at mitigating bias by considering the causal relationships among variables and incorporating them into the reweighting process. The proposed method adopts two neural networks, whose structures are intentionally used to reflect the structures of a causal graph and of an interventional graph. The two neural networks can approximate the causal model of the data, and the causal model of interventions. Furthermore, reweighting guided by a discriminator is applied to achieve various fairness notions. Experiments on real-world datasets show that our method can achieve causal fairness on the data while remaining close to the original data for downstream tasks.

cs.LG↗
arXiv2023

Convergence analysis of variable steps BDF2 method for the space fractional Cahn-Hilliard model

An implicit variable-step BDF2 scheme is established for solving the space fractional Cahn-Hilliard equation, involving the fractional Laplacian, derived from a gradient flow in the negative order Sobolev space $H^{-α}$, $α\in(0,1)$. The Fourier pseudo-spectral method is applied for the spatial approximation. The proposed scheme inherits the energy dissipation law in the form of the modified discrete energy under the sufficient restriction of the time-step ratios. The convergence of the fully discrete scheme is rigorously provided utilizing the newly proved discrete embedding type convolution inequality dealing with the fractional Laplacian. Besides, the mass conservation and the unique solvability are also theoretically guaranteed. Numerical experiments are carried out to show the accuracy and the energy dissipation both for various interface widths. In particular, the multiple-time-scale evolution of the solution is captured by an adaptive time-stepping strategy in the short-to-long time simulation.

math.NA↗
arXiv2023

High-Fidelity Image Synthesis from Pulmonary Nodule Lesion Maps using Semantic Diffusion Model

Lung cancer has been one of the leading causes of cancer-related deaths worldwide for years. With the emergence of deep learning, computer-assisted diagnosis (CAD) models based on learning algorithms can accelerate the nodule screening process, providing valuable assistance to radiologists in their daily clinical workflows. However, developing such robust and accurate models often requires large-scale and diverse medical datasets with high-quality annotations. Generating synthetic data provides a pathway for augmenting datasets at a larger scale. Therefore, in this paper, we explore the use of Semantic Diffusion Mod- els (SDM) to generate high-fidelity pulmonary CT images from segmentation maps. We utilize annotation information from the LUNA16 dataset to create paired CT images and masks, and assess the quality of the generated images using the Frechet Inception Distance (FID), as well as on two common clinical downstream tasks: nodule detection and nodule localization. Achieving improvements of 3.96% for detection accuracy and 8.50% for AP50 in nodule localization task, respectively, demonstrates the feasibility of the approach.

eess.IV↗
arXiv2023

Pressure tuning of optical reflectivity in LuH2

Recently, the claim of room-temperature superconductivity in nitrogen-doped lutetium hydride at near-ambient conditions has attracted tremendous attention. Criticism of the work rises shortly, while further explorations are needed to settle the dispute. One of the intriguing observations is the pressured-induced color change, which has been reproduced in the lutetium dihydride LuH2 while its mechanism remains unclear. Through optical reflectivity measurements of LuH2 in the visible to near-infrared region, we observe strong light absorption next to the sharp plasmon resonance, which continuously shifts to higher energies with increasing pressure. It gives rise to the increased reflection of red light and suppressed reflection of blue light. Our work sheds light on resolving the puzzles regarding the pressure induced color change in LuH2.

cond-mat.supr-con↗
arXiv2023

Observation of Rydberg moiré excitons

Rydberg excitons, the solid-state counterparts of Rydberg atoms, have sparked considerable interest in harnessing their quantum application potentials, whereas a major challenge is realizing their spatial confinement and manipulation. Lately, the rise of two-dimensional moiré superlattices with highly tunable periodic potentials provides a possible pathway. Here, we experimentally demonstrate this capability through the observation of Rydberg moiré excitons (XRM), which are moiré trapped Rydberg excitons in monolayer semiconductor WSe2 adjacent to twisted bilayer graphene. In the strong coupling regime, the XRM manifest as multiple energy splittings, pronounced redshift, and narrowed linewidth in the reflectance spectra, highlighting their charge-transfer character where electron-hole separation is enforced by the strongly asymmetric interlayer Coulomb interactions. Our findings pave the way for pursuing novel physics and quantum technology exploitation based on the excitonic Rydberg states.

cond-mat.mes-hall↗
arXiv2023

Energy stability and convergence of variable-step L1 scheme for the time fractional Swift-Hohenberg model

A fully implicit numerical scheme is established for solving the time fractional Swift-Hohenberg (TFSH) equation with a Caputo time derivative of order $α\in(0,1)$. The variable-step L1 formula and the finite difference method are employed for the time and the space discretizations, respectively. The unique solvability of the numerical scheme is proved by the Brouwer fixed-point theorem. With the help of the discrete convolution form of L1 formula, the time-stepping scheme is shown to preserve a discrete energy dissipation law which is asymptotically compatible with the classic energy law as $α\to1^-$. Furthermore, the $L^\infty$ norm boundedness of the discrete solution is obtained. Combining with the global consistency error analysis framework, the $L^2$ norm convergence order is shown rigorously. Several numerical examples are provided to illustrate the accuracy and the energy dissipation law of the proposed method. In particular, the adaptive time-stepping strategy is utilized to capture the multi-scale time behavior of the TFSH model efficiently.

math.NA↗
arXiv2023

Energy stable and $L^2$ norm convergent BDF3 scheme for the Swift-Hohenberg equation

A fully discrete implicit scheme is proposed for the Swift-Hohenberg model, combining the third-order backward differentiation formula (BDF3) for the time discretization and the second-order finite difference scheme for the space discretization. Applying the Brouwer fixed-point theorem and the positive definiteness of the convolution coefficients of BDF3, the presented numerical algorithm is proved to be uniquely solvable and unconditionally energy stable, further, the numerical solution is shown to be bounded in the maximum norm. The proposed scheme is rigorously proved to be convergent in $L^2$ norm by the discrete orthogonal convolution (DOC) kernel, which transfer the four-level-solution form into the three-level-gradient form for the approximation of the temporal derivative. Consequently, the error estimate for the numerical solution is established by utilization of the discrete Gronwall inequality. Numerical examples in 2D and 3D cases are provided to support the theoretical results.

math.NA↗