arXiv subjects

Xiao Hu

Publications and source records attributed to Xiao Hu.

At least 127 records · Page 7Linked to original sources

arXiv2025

The influence of Structural Dynamics in Two-Dimensional Hybrid Organic-Inorganic Perovskites on their Photoluminescence Efficiency -- Neutron scattering analysis

Two-dimensional hybrid organic-inorganic perovskites (HOIPs) have emerged as promising materials for light-emitting diode applications. In this study, by using time-of-flight neutron spectroscopy we identified and quantitatively separated the lattice vibrational and molecular rotational dynamics of two perovskites, butylammonium lead iodide (BA)$_{2}$PbI$_{4}$ and phenethyl-ammonium lead iodide (PEA)$_{2}$PbI$_{4}$. By examining the corresponding temperature dependence, we found that the lattice vibrations, as evidenced by neutron spectra, are consistent with the lattice dynamics obtained from Raman scattering. We revealed that the rotational dynamics of organic molecules in these materials tend to suppress their photoluminescence quantum yield (PLQY) while the vibrational dynamics did not show predominant correlations with the same. Additionally, we observed photoluminescence emission peak splitting for both systems, which becomes prominent above certain critical temperatures where the suppression of PLQY begins. This study suggests that the rotational motions of polarized molecules may lead to a reduction in exciton binding energy or the breaking of degeneracy in exciton binding energy levels, enhancing non-radiative recombination rates, and consequently reducing photoluminescence yield. These findings offer a deeper understanding of fundamental interactions in 2D HOIPs and could guide the design of more efficient light-emitting materials for advanced technological applications.

cond-mat.mtrl-sci↗

arXiv2025

MARAGE: Transferable Multi-Model Adversarial Attack for Retrieval-Augmented Generation Data Extraction

Retrieval-Augmented Generation (RAG) offers a solution to mitigate hallucinations in Large Language Models (LLMs) by grounding their outputs to knowledge retrieved from external sources. The use of private resources and data in constructing these external data stores can expose them to risks of extraction attacks, in which attackers attempt to steal data from these private databases. Existing RAG extraction attacks often rely on manually crafted prompts, which limit their effectiveness. In this paper, we introduce a framework called MARAGE for optimizing an adversarial string that, when appended to user queries submitted to a target RAG system, causes outputs containing the retrieved RAG data verbatim. MARAGE leverages a continuous optimization scheme that integrates gradients from multiple models with different architectures simultaneously to enhance the transferability of the optimized string to unseen models. Additionally, we propose a strategy that emphasizes the initial tokens in the target RAG data, further improving the attack's generalizability. Evaluations show that MARAGE consistently outperforms both manual and optimization-based baselines across multiple LLMs and RAG datasets, while maintaining robust transferability to previously unseen models. Moreover, we conduct probing tasks to shed light on the reasons why MARAGE is more effective compared to the baselines and to analyze the impact of our approach on the model's internal state.

cs.CL↗

arXiv2025

Boundedly finite-to-one functions

A function is boundedly finite-to-one if there is a natural number $k$ such that each point has at most $k$ inverse images. In this paper, we prove in $\mathsf{ZF}$ (i.e., the Zermelo--Fraenkel set theory without the axiom of choice) several results concerning this notion, among which are the following: (1) For each infinite set $A$ and natural number $n$, there is no boundedly finite-to-one function from $\mathcal{S}(A)$ to $\mathcal{S}_{\leq n}(A)$, where $\mathcal{S}(A)$ is the set of all permutations of $A$ and $\mathcal{S}_{\leq n}(A)$ is the set of all permutations of $A$ moving at most $n$ points. (2) For each infinite set $A$, there is no boundedly finite-to-one function from $\mathcal{B}(A)$ to $\mathrm{fin}(A)$, where $\mathcal{B}(A)$ is the set of all partitions of $A$ such that every block is finite and $\mathrm{fin}(A)$ is the set of all finite subsets of $A$.

math.LO↗

arXiv2025

Mining Social Determinants of Health for Heart Failure Patient 30-Day Readmission via Large Language Model

Heart Failure (HF) affects millions of Americans and leads to high readmission rates, posing significant healthcare challenges. While Social Determinants of Health (SDOH) such as socioeconomic status and housing stability play critical roles in health outcomes, they are often underrepresented in structured EHRs and hidden in unstructured clinical notes. This study leverages advanced large language models (LLMs) to extract SDOHs from clinical text and uses logistic regression to analyze their association with HF readmissions. By identifying key SDOHs (e.g. tobacco usage, limited transportation) linked to readmission risk, this work also offers actionable insights for reducing readmissions and improving patient care.

cs.LG↗

arXiv2025

Optimal Oblivious Algorithms for Multi-way Joins

In cloud databases, cloud computation over sensitive data uploaded by clients inevitably causes concern about data security and privacy. Even when encryption primitives and trusted computing environments are integrated into query processing to safeguard the actual contents of the data, access patterns of algorithms can still leak private information about the data. Oblivious Random Access Memory (ORAM) and circuits are two generic approaches to address this issue, ensuring that access patterns of algorithms remain oblivious to the data. However, deploying these methods on insecure algorithms, particularly for multi-way join processing, is computationally expensive and inherently challenging. In this paper, we propose a novel sorting-based algorithm for multi-way join processing that operates without relying on ORAM simulations or other security assumptions. Our algorithm is a non-trivial, provably oblivious composition of basic primitives, with time complexity matching the insecure worst-case optimal join algorithm, up to a logarithmic factor. Furthermore, it is cache-agnostic, with cache complexity matching the insecure lower bound, also up to a logarithmic factor. This clean and straightforward approach has the potential to be extended to other security settings and implemented in practical database systems.

cs.DB↗

arXiv2024

3D Gap Opening in Non-Ideal MHD Protoplanetary Disks: Asymmetric Accretion, Meridional Vortices, and Observational Signatures

Recent high-angular resolution ALMA observations have revealed rich information about protoplanetary disks, including ubiquitous substructures and three-dimensional gas kinematics at different emission layers. One interpretation of these observations is embedded planets. Previous 3-D planet-disk interaction studies are either based on viscous simulations, or non-ideal magnetohydrodynamics (MHD) simulations with simple prescribed magnetic diffusivities. This study investigates the dynamics of gap formation in 3-D non-ideal MHD disks using non-ideal MHD coefficients from the look-up table that is self-consistently calculated based on the thermo-chemical code. We find a concentration of the poloidal magnetic flux in the planet-opened gap (in agreement with previous work) and enhanced field-matter coupling due to gas depletion, which together enable efficient magnetic braking of the gap material, driving a fast accretion layer significantly displaced from the disk midplane. The fast accretion helps deplete the gap further and is expected to negatively impact the growth of planetary embryos. It also affects the corotation torque by shrinking the region of horseshoe orbits on the trailing side of the planet. Together with the magnetically driven disk wind, the fast accretion layer generates a large, persistent meridional vortex in the gap, which breaks the mirror symmetry of gas kinematics between the top and bottom disk surfaces. Finally, by studying the kinematics at the emission surfaces, we discuss the implications of planets in realistic non-ideal MHD disks on kinematics observations.

astro-ph.EP↗

arXiv2024

A Visual Cooperative Localization Method for Airborne Magnetic Surveying Based on a Manifold Sensor Fusion Algorithm Using Lie Groups

Recent advancements in UAV technology have spurred interest in developing multi-UAV aerial surveying systems for use in confined environments where GNSS signals are blocked or jammed. This paper focuses airborne magnetic surveying scenarios. To obtain clean magnetic measurements reflecting the Earth's magnetic field, the magnetic sensor must be isolated from other electronic devices, creating a significant localization challenge. We propose a visual cooperative localization solution. The solution incorporates a visual processing module and an improved manifold-based sensor fusion algorithm, delivering reliable and accurate positioning information. Real flight experiments validate the approach, demonstrating single-axis centimeter-level accuracy and decimeter-level overall 3D positioning accuracy.

eess.SP↗

arXiv2024

Sparse learned kernels for interpretable and efficient medical time series processing

Rapid, reliable, and accurate interpretation of medical time-series signals is crucial for high-stakes clinical decision-making. Deep learning methods offered unprecedented performance in medical signal processing but at a cost: they were compute-intensive and lacked interpretability. We propose Sparse Mixture of Learned Kernels (SMoLK), an interpretable architecture for medical time series processing. SMoLK learns a set of lightweight flexible kernels that form a single-layer sparse neural network, providing not only interpretability, but also efficiency, robustness, and generalization to unseen data distributions. We introduce a parameter reduction techniques to reduce the size of SMoLK's networks while maintaining performance. We test SMoLK on two important tasks common to many consumer wearables: photoplethysmography (PPG) artifact detection and atrial fibrillation detection from single-lead electrocardiograms (ECGs). We find that SMoLK matches the performance of models orders of magnitude larger. It is particularly suited for real-time applications using low-power devices, and its interpretability benefits high-stakes situations.

eess.SP↗

arXiv2024

Rossby Wave Instability and Substructure Formation in 3D Non-Ideal MHD Wind-Launching Disks

Rings and gaps are routinely observed in the dust continuum emission of protoplanetary discs (PPDs). How they form and evolve remains debated. Previous studies have demonstrated the possibility of spontaneous gas rings and gaps formation in wind-launching disks. Here, we show that such gas substructures are unstable to the Rossby Wave Instability (RWI) through numerical simulations. Specifically, shorter wavelength azimuthal modes develop earlier, and longer wavelength ones dominate later, forming elongated (arc-like) anti-cyclonic vortices in the rings and (strongly magnetized) cyclonic vortices in the gaps that persist until the end of the simulation. Highly elongated vortices with aspect ratios of 10 or more are found to decay with time in our non-ideal MHD simulation, in contrast with the hydro case. This difference could be caused by magnetically induced motions, particularly strong meridional circulations with large values of the azimuthal component of the vorticity, which may be incompatible with the columnar structure preferred by vortices. The cyclonic and anti-cyclonic RWI vortices saturate at moderate levels, modifying but not destroying the rings and gaps in the radial gas distribution of the disk. In particular, they do not shut off the poloidal magnetic flux accumulation in low-density regions and the characteristic meridional flow patterns that are crucial to the ring and gap formation in wind-launching disks. Nevertheless, the RWI and their associated vortices open up the possibility of producing non-axisymmetric dust features observed in a small fraction of protoplanetary disks through non-ideal MHD, although detailed dust treatment is needed to explore this possibility.

astro-ph.EP↗

arXiv2024

Comment on "Absence of Topological Protection of the Interface States in $\mathbb{Z}_2$ Photonic Crystals"

In the Letter, Xu et al. reported that edge modes disappear in the expanded structure of Wu-Hu model characterized by $\mathbb{Z}_2$ topological index, while appear in the trivial shrunken structure, when the edge cuts through the hexagonal unit cell. They then concluded that these edge modes are defect modes lacking topological protection. Unfortunately, their approach is not justified, rendering the conclusion unsolid.

physics.optics↗

arXiv2024

Query-Policy Misalignment in Preference-Based Reinforcement Learning

Preference-based reinforcement learning (PbRL) provides a natural way to align RL agents' behavior with human desired outcomes, but is often restrained by costly human feedback. To improve feedback efficiency, most existing PbRL methods focus on selecting queries to maximally improve the overall quality of the reward model, but counter-intuitively, we find that this may not necessarily lead to improved performance. To unravel this mystery, we identify a long-neglected issue in the query selection schemes of existing PbRL studies: Query-Policy Misalignment. We show that the seemingly informative queries selected to improve the overall quality of reward model actually may not align with RL agents' interests, thus offering little help on policy learning and eventually resulting in poor feedback efficiency. We show that this issue can be effectively addressed via near on-policy query and a specially designed hybrid experience replay, which together enforce the bidirectional query-policy alignment. Simple yet elegant, our method can be easily incorporated into existing approaches by changing only a few lines of code. We showcase in comprehensive experiments that our method achieves substantial gains in both human feedback and RL sample efficiency, demonstrating the importance of addressing query-policy misalignment in PbRL tasks.

cs.LG↗

arXiv2024

Adapting Job Recommendations to User Preference Drift with Behavioral-Semantic Fusion Learning

Job recommender systems are crucial for aligning job opportunities with job-seekers in online job-seeking. However, users tend to adjust their job preferences to secure employment opportunities continually, which limits the performance of job recommendations. The inherent frequency of preference drift poses a challenge to promptly and precisely capture user preferences. To address this issue, we propose a novel session-based framework, BISTRO, to timely model user preference through fusion learning of semantic and behavioral information. Specifically, BISTRO is composed of three stages: 1) coarse-grained semantic clustering, 2) fine-grained job preference extraction, and 3) personalized top-$k$ job recommendation. Initially, BISTRO segments the user interaction sequence into sessions and leverages session-based semantic clustering to achieve broad identification of person-job matching. Subsequently, we design a hypergraph wavelet learning method to capture the nuanced job preference drift. To mitigate the effect of noise in interactions caused by frequent preference drift, we innovatively propose an adaptive wavelet filtering technique to remove noisy interaction. Finally, a recurrent neural network is utilized to analyze session-based interaction for inferring personalized preferences. Extensive experiments on three real-world offline recruitment datasets demonstrate the significant performances of our framework. Significantly, BISTRO also excels in online experiments, affirming its effectiveness in live recruitment settings. This dual success underscores the robustness and adaptability of BISTRO. The source code is available at https://github.com/Applied-Machine-Learning-Lab/BISTRO.

cs.IR↗

arXiv2024

Spin waves in Dirac semimetal Ca$_{0.6}$Sr$_{0.4}$MnSb$_2$ investigated with neutrons by the diffraction method

We report neutron diffraction measurements of Ca$_{0.6}$Sr$_{0.4}$MnSb$_2$, a low-carrier-density Dirac semimetal in which the antiferromagnetic Mn layers are interleaved with Sb layers that host Dirac fermions. We have discovered that we can detect a good quality inelastic spin wave signal from a small (m ~ 0.28 g) single crystal sample by the diffraction method, without energy analysis, using a neutron diffractometer with a position-sensitive area detector; the spin-waves appear as diffuse scattering that is shaped by energy-momentum conservation. By fitting this characteristic magnetic scattering to a spin-wave model, we refine all parameters of the model spin Hamiltonian, including the inter-plane interaction, through use of a three-dimensional measurement in reciprocal space. We also measure the temperature dependence of the spin waves, including the softening of the spin gap on approaching the Neel temperature, $T_N$. Not only do our results provide important new insights into an interplay of magnetism and Dirac electrons, they also establish a new, high-throughput approach to characterizing magnetic excitations on a modern diffractometer without direct energy analysis. Our work opens exciting new opportunities for the follow-up parametric and compositional studies on small, ~0.1 g crystals.

cond-mat.str-el↗

arXiv2024

From Basic to Extra Features: Hypergraph Transformer Pretrain-then-Finetuning for Balanced Clinical Predictions on EHR

Electronic Health Records (EHRs) contain rich patient information and are crucial for clinical research and practice. In recent years, deep learning models have been applied to EHRs, but they often rely on massive features, which may not be readily available for all patients. We propose HTP-Star, which leverages hypergraph structures with a pretrain-then-finetune framework for modeling EHR data, enabling seamless integration of additional features. Additionally, we design two techniques, namely (1) Smoothness-inducing Regularization and (2) Group-balanced Reweighting, to enhance the model's robustness during fine-tuning. Through experiments conducted on two real EHR datasets, we demonstrate that HTP-Star consistently outperforms various baselines while striking a balance between patients with basic and extra features.

cs.LG↗

arXiv2024

DecisionNCE: Embodied Multimodal Representations via Implicit Preference Learning

Multimodal pretraining is an effective strategy for the trinity of goals of representation learning in autonomous robots: 1) extracting both local and global task progressions; 2) enforcing temporal consistency of visual representation; 3) capturing trajectory-level language grounding. Most existing methods approach these via separate objectives, which often reach sub-optimal solutions. In this paper, we propose a universal unified objective that can simultaneously extract meaningful task progression information from image sequences and seamlessly align them with language instructions. We discover that via implicit preferences, where a visual trajectory inherently aligns better with its corresponding language instruction than mismatched pairs, the popular Bradley-Terry model can transform into representation learning through proper reward reparameterizations. The resulted framework, DecisionNCE, mirrors an InfoNCE-style objective but is distinctively tailored for decision-making tasks, providing an embodied representation learning framework that elegantly extracts both local and global task progression features, with temporal consistency enforced through implicit time contrastive learning, while ensuring trajectory-level instruction grounding via multimodal joint encoding. Evaluation on both simulated and real robots demonstrates that DecisionNCE effectively facilitates diverse downstream policy learning tasks, offering a versatile solution for unified representation and reward learning. Project Page: https://2toinf.github.io/DecisionNCE/

cs.RO↗

arXiv2024

PromptLink: Leveraging Large Language Models for Cross-Source Biomedical Concept Linking

Linking (aligning) biomedical concepts across diverse data sources enables various integrative analyses, but it is challenging due to the discrepancies in concept naming conventions. Various strategies have been developed to overcome this challenge, such as those based on string-matching rules, manually crafted thesauri, and machine learning models. However, these methods are constrained by limited prior biomedical knowledge and can hardly generalize beyond the limited amounts of rules, thesauri, or training samples. Recently, large language models (LLMs) have exhibited impressive results in diverse biomedical NLP tasks due to their unprecedentedly rich prior knowledge and strong zero-shot prediction abilities. However, LLMs suffer from issues including high costs, limited context length, and unreliable predictions. In this research, we propose PromptLink, a novel biomedical concept linking framework that leverages LLMs. It first employs a biomedical-specialized pre-trained language model to generate candidate concepts that can fit in the LLM context windows. Then it utilizes an LLM to link concepts through two-stage prompts, where the first-stage prompt aims to elicit the biomedical prior knowledge from the LLM for the concept linking task and the second-stage prompt enforces the LLM to reflect on its own predictions to further enhance their reliability. Empirical results on the concept linking task between two EHR datasets and an external biomedical KG demonstrate the effectiveness of PromptLink. Furthermore, PromptLink is a generic framework without reliance on additional prior knowledge, context, or training data, making it well-suited for concept linking across various types of data sources. The source code is available at https://github.com/constantjxyz/PromptLink.

cs.IR↗

arXiv2024

Guidance Design for Escape Flight Vehicle Using Evolution Strategy Enhanced Deep Reinforcement Learning

Guidance commands of flight vehicles are a series of data sets with fixed time intervals, thus guidance design constitutes a sequential decision problem and satisfies the basic conditions for using deep reinforcement learning (DRL). In this paper, we consider the scenario where the escape flight vehicle (EFV) generates guidance commands based on DRL and the pursuit flight vehicle (PFV) generates guidance commands based on the proportional navigation method. For the EFV, the objective of the guidance design entails progressively maximizing the residual velocity, subject to the constraint imposed by the given evasion distance. Thus an irregular dynamic max-min problem of extremely large-scale is formulated, where the time instant when the optimal solution can be attained is uncertain and the optimum solution depends on all the intermediate guidance commands generated before. For solving this problem, a two-step strategy is conceived. In the first step, we use the proximal policy optimization (PPO) algorithm to generate the guidance commands of the EFV. The results obtained by PPO in the global search space are coarse, despite the fact that the reward function, the neural network parameters and the learning rate are designed elaborately. Therefore, in the second step, we propose to invoke the evolution strategy (ES) based algorithm, which uses the result of PPO as the initial value, to further improve the quality of the solution by searching in the local space. Simulation results demonstrate that the proposed guidance design method based on the PPO algorithm is capable of achieving a residual velocity of 67.24 m/s, higher than the residual velocities achieved by the benchmark soft actor-critic and deep deterministic policy gradient algorithms. Furthermore, the proposed ES-enhanced PPO algorithm outperforms the PPO algorithm by 2.7\%, achieving a residual velocity of 69.04 m/s.

cs.LG↗

arXiv2024

SiamQuality: A ConvNet-Based Foundation Model for Imperfect Physiological Signals

Foundation models, especially those using transformers as backbones, have gained significant popularity, particularly in language and language-vision tasks. However, large foundation models are typically trained on high-quality data, which poses a significant challenge, given the prevalence of poor-quality real-world data. This challenge is more pronounced for developing foundation models for physiological data; such data are often noisy, incomplete, or inconsistent. The present work aims to provide a toolset for developing foundation models on physiological data. We leverage a large dataset of photoplethysmography (PPG) signals from hospitalized intensive care patients. For this data, we propose SimQuality, a novel self-supervised learning task based on convolutional neural networks (CNNs) as the backbone to enforce representations to be similar for good and poor quality signals that are from similar physiological states. We pre-trained the SimQuality on over 36 million 30-second PPG pairs and then fine-tuned and tested on six downstream tasks using external datasets. The results demonstrate the superiority of the proposed approach on all the downstream tasks, which are extremely important for heart monitoring on wearable devices. Our method indicates that CNNs can be an effective backbone for foundation models that are robust to training data quality.

eess.SP↗