arXiv subjects

Xin Zheng

Publications and source records attributed to Xin Zheng.

At least 55 records · Page 3Linked to original sources

arXiv2025

Direct measurement of the 3P0 clock state natural lifetime in 87Sr

Optical lattice clocks based on the narrow (5s2)1S0 - (5s5p)3P0 transition in neutral strontium (Sr) are among the most precise and accurate measurement devices in existence. Although this transition is completely forbidden by selection rules, state mixing from the hyperfine interaction in 87Sr provides a weakly allowed transition that can be coherently driven with practical clock laser intensities. While the coherent interrogation times of optical clocks are typically set by the linewidth of the probe laser, this limitation can be overcome in synchronous differential comparisons between ensembles. In such measurements the natural lifetime of the 1S0-3P0 clock transition becomes the fundamental limiting factor to the duration of a single run of the experiment. However, a direct measurement of the decay rate of the clock excited state is quite challenging due to the competing effects of other loss channels such as Raman scattering, inelastic collisions and atom-loss due to background gas. In this work, we monitor the decay of Sr atoms trapped in an optical lattice and initialized in the 3P0 state. By making measurements of high and low density ensembles of both 87Sr and 88Sr across varying lattice trap depths, we isolate radiative decay, which accounts for a significant fraction of the observed decays at low depths. We obtain a natural radiative decay lifetime of 167(+79/-40) s for the 3P0 clock state in 87Sr, a value that is consistent with previously reported measurements and theoretical predictions. We also introduce an additional measurement scheme that involves repeated measurements of the ground state population within a single experimental sequence, validating our model and the consistency of the measured rates.

physics.atom-ph↗

arXiv2025

Critic-CoT: Boosting the reasoning abilities of large language model via Chain-of-thoughts Critic

Self-critic has become a crucial mechanism for enhancing the reasoning performance of LLMs. However, current approaches mainly involve basic prompts for intuitive instance-level feedback, which resembles System-1 processes and limits the reasoning capabilities. Moreover, there is a lack of in-depth investigations into the relationship between LLM's ability to criticize and its task-solving performance. To address these issues, we propose Critic-CoT, a novel framework that pushes LLMs toward System-2-like critic capability. Through a step-wise CoT reasoning paradigm and the automatic construction of distant-supervision data without human annotation, Critic-CoT enables LLMs to engage in slow, analytic self-critique and refinement, thereby improving their reasoning abilities. Experiments on GSM8K and MATH demonstrate that our enhanced model significantly boosts task-solving performance by filtering out invalid solutions or iterative refinement. Furthermore, we investigate the intrinsic correlation between critique and task-solving abilities within LLMs, discovering that these abilities can mutually reinforce each other rather than conflict.

cs.CL↗

arXiv2025

Adaptive Sentencing Prediction with Guaranteed Accuracy and Legal Interpretability

Existing research on judicial sentencing prediction predominantly relies on end-to-end models, which often neglect the inherent sentencing logic and lack interpretability-a critical requirement for both scholarly research and judicial practice. To address this challenge, we make three key contributions:First, we propose a novel Saturated Mechanistic Sentencing (SMS) model, which provides inherent legal interpretability by virtue of its foundation in China's Criminal Law. We also introduce the corresponding Momentum Least Mean Squares (MLMS) adaptive algorithm for this model. Second, for the MLMS algorithm based adaptive sentencing predictor, we establish a mathematical theory on the accuracy of adaptive prediction without resorting to any stationarity and independence assumptions on the data. We also provide a best possible upper bound for the prediction accuracy achievable by the best predictor designed in the known parameters case. Third, we construct a Chinese Intentional Bodily Harm (CIBH) dataset. Utilizing this real-world data, extensive experiments demonstrate that our approach achieves a prediction accuracy that is not far from the best possible theoretical upper bound, validating both the model's suitability and the algorithm's accuracy.

cs.LG↗

arXiv2025

Unconventional band splitting of CeSb in the devil's staircase transition

The interplay between magnetism and electronic band structure is a central theme in condensed matter physics. CeSb, with its complex devil's staircase antiferromagnetic transition, offers a unique opportunity to explore this interplay. Using angle-resolved photoemission spectroscopy (ARPES), we investigate the electronic structure evolution across the devil's staircase transition. Upon entering the antiferromagnetic phase, we observe an intriguing band splitting of the electron pocket around the X point. The energy separation between the split bands changes abruptly with temperature, consistent with the characteristics of the first-order phase transition. However, their respective spectral weights behave gradually with temperature. Combined with our density functional theory (DFT) calculations, we suggest that this atypical behavior deviates from conventional magnetically induced band splitting and potentially arises from the intricate modulation of paramagnetic and antiferromagnetic layers within the devil's staircase transition. Our results provide insights into the complex relationship between electronic structure and magnetism in correlated electron systems.

cond-mat.str-el↗

arXiv2025

FedSA: A Unified Representation Learning via Semantic Anchors for Prototype-based Federated Learning

Prototype-based federated learning has emerged as a promising approach that shares lightweight prototypes to transfer knowledge among clients with data heterogeneity in a model-agnostic manner. However, existing methods often collect prototypes directly from local models, which inevitably introduce inconsistencies into representation learning due to the biased data distributions and differing model architectures among clients. In this paper, we identify that both statistical and model heterogeneity create a vicious cycle of representation inconsistency, classifier divergence, and skewed prototype alignment, which negatively impacts the performance of clients. To break the vicious cycle, we propose a novel framework named Federated Learning via Semantic Anchors (FedSA) to decouple the generation of prototypes from local representation learning. We introduce a novel perspective that uses simple yet effective semantic anchors serving as prototypes to guide local models in learning consistent representations. By incorporating semantic anchors, we further propose anchor-based regularization with margin-enhanced contrastive learning and anchor-based classifier calibration to correct feature extractors and calibrate classifiers across clients, achieving intra-class compactness and inter-class separability of prototypes while ensuring consistent decision boundaries. We then update the semantic anchors with these consistent and discriminative prototypes, which iteratively encourage clients to collaboratively learn a unified data representation with robust generalization. Extensive experiments under both statistical and model heterogeneity settings show that FedSA significantly outperforms existing prototype-based FL methods on various classification tasks.

cs.LG↗

arXiv2025

ORLM: A Customizable Framework in Training Large Models for Automated Optimization Modeling

Optimization modeling plays a critical role in the application of Operations Research (OR) tools to address real-world problems, yet they pose challenges and require extensive expertise from OR experts. With the advent of large language models (LLMs), new opportunities have emerged to streamline and automate such task. However, current research predominantly relies on closed-source LLMs such as GPT-4, along with extensive prompt engineering techniques. This reliance stems from the scarcity of high-quality training datasets for optimization modeling, resulting in elevated costs, prolonged processing times, and privacy concerns. To address these challenges, our work is the first to propose a viable path for training open-source LLMs that are capable of optimization modeling and developing solver codes, eventually leading to a superior ability for automating optimization modeling and solving. Particularly, we design the {\sc OR-Instruct}, a semi-automated data synthesis framework for optimization modeling that enables customizable enhancements for specific scenarios or model types. This work also introduces IndustryOR, the first industrial benchmark for evaluating LLMs in solving practical OR problems. We train several 7B-scale open-source LLMs using synthesized data (dubbed ORLMs{https://github.com/Cardinal-Operations/ORLM}), which exhibit significantly enhanced optimization modeling capabilities, achieving competitive performance across the NL4OPT, MAMO, and IndustryOR benchmarks. Additionally, our experiments highlight the potential of scaling law and reinforcement learning to further enhance the performance of ORLMs. The workflows and human-machine interaction paradigms of ORLMs in practical industrial applications are also discussed in the paper.

cs.CL↗

arXiv2025

A Label-Free Heterophily-Guided Approach for Unsupervised Graph Fraud Detection

Graph fraud detection (GFD) has rapidly advanced in protecting online services by identifying malicious fraudsters. Recent supervised GFD research highlights that heterophilic connections between fraudsters and users can greatly impact detection performance, since fraudsters tend to camouflage themselves by building more connections to benign users. Despite the promising performance of supervised GFD methods, the reliance on labels limits their applications to unsupervised scenarios; Additionally, accurately capturing complex and diverse heterophily patterns without labels poses a further challenge. To fill the gap, we propose a Heterophily-guided Unsupervised Graph fraud dEtection approach (HUGE) for unsupervised GFD, which contains two essential components: a heterophily estimation module and an alignment-based fraud detection module. In the heterophily estimation module, we design a novel label-free heterophily metric called HALO, which captures the critical graph properties for GFD, enabling its outstanding ability to estimate heterophily from node attributes. In the alignment-based fraud detection module, we develop a joint MLP-GNN architecture with ranking loss and asymmetric alignment loss. The ranking loss aligns the predicted fraud score with the relative order of HALO, providing an extra robustness guarantee by comparing heterophily among non-adjacent nodes. Moreover, the asymmetric alignment loss effectively utilizes structural information while alleviating the feature-smooth effects of GNNs. Extensive experiments on 6 datasets demonstrate that HUGE significantly outperforms competitors, showcasing its effectiveness and robustness.

cs.LG↗

arXiv2024

Brain-to-Text Benchmark '24: Lessons Learned

Speech brain-computer interfaces aim to decipher what a person is trying to say from neural activity alone, restoring communication to people with paralysis who have lost the ability to speak intelligibly. The Brain-to-Text Benchmark '24 and associated competition was created to foster the advancement of decoding algorithms that convert neural activity to text. Here, we summarize the lessons learned from the competition ending on June 1, 2024 (the top 4 entrants also presented their experiences in a recorded webinar). The largest improvements in accuracy were achieved using an ensembling approach, where the output of multiple independent decoders was merged using a fine-tuned large language model (an approach used by all 3 top entrants). Performance gains were also found by improving how the baseline recurrent neural network (RNN) model was trained, including by optimizing learning rate scheduling and by using a diphone training objective. Improving upon the model architecture itself proved more difficult, however, with attempts to use deep state space models or transformers not yet appearing to offer a benefit over the RNN baseline. The benchmark will remain open indefinitely to support further work towards increasing the accuracy of brain-to-text algorithms.

cs.CL↗

arXiv2024

$L_1$-Based Adaptive Identification with Saturated Observations

It is well-known that saturated output observations are prevalent in various practical systems and that the $\ell_1$-norm is more robust than the $\ell_2$-norm-based parameter estimation. Unfortunately, adaptive identification based on both saturated observations and the $\ell_1$-optimization turns out to be a challenging nonlinear problem, and has rarely been explored in the literature. Motivated by this and the need to fit with the $\ell_1$-based index of prediction accuracy in, e.g., judicial sentencing prediction problems, we propose a two-step weighted $\ell_1$-based adaptive identification algorithm. Under certain excitation conditions much weaker than the traditional persistent excitation (PE) condition, we will establish the global convergence of both the parameter estimators and the adaptive predictors. It is worth noting that our results do not rely on the widely used independent and identically distributed (iid) assumptions on the system signals, and thus do not exclude applications to feedback control systems. We will demonstrate the advantages of our proposed new adaptive algorithm over the existing $\ell_2$-based ones, through both a numerical example and a real-data-based sentencing prediction problem.

eess.SY↗

arXiv2024

Local convergence analysis of L1/finite element scheme for a constant delay reaction-subdiffusion equation with uniform time mesh

The aim of this paper is to develop a refined error estimate of L1/finite element scheme for a reaction-subdiffusion equation with constant delay $τ$ and uniform time mesh. Under the non-uniform multi-singularity assumption of exact solution in time, the local truncation errors of the L1 scheme with uniform mesh is investigated. Then we introduce a fully discrete finite element scheme of the considered problem. Next, a novel discrete fractional Grönwall inequality with constant delay term is proposed, which does not include the increasing Mittag-Leffler function comparing with some popular other cases. By applying this Grönwall inequality, we obtain the pointwise-in-time and piecewise-in-time error estimates of the finite element scheme without the Mittag-Leffler function. In particular, the latter shows that, for the considered interval $((i-1)τ,iτ]$, although the convergence in time is low for $i=1$, it will be improved as the increasing $i$, which is consistent with the factual assumption that the smoothness of the solution will be improved as the increasing $i$. Finally, we present some numerical tests to verify the developed theory.

math.NA↗

arXiv2024

SIAM: A Simple Alternating Mixer for Video Prediction

Video prediction, predicting future frames from the previous ones, has broad applications such as autonomous driving and weather forecasting. Existing state-of-the-art methods typically focus on extracting either spatial, temporal, or spatiotemporal features from videos. Different feature focuses, resulting from different network architectures, may make the resultant models excel at some video prediction tasks but perform poorly on others. Towards a more generic video prediction solution, we explicitly model these features in a unified encoder-decoder framework and propose a novel simple alternating Mixer (SIAM). The novelty of SIAM lies in the design of dimension alternating mixing (DaMi) blocks, which can model spatial, temporal, and spatiotemporal features through alternating the dimensions of the feature maps. Extensive experimental results demonstrate the superior performance of the proposed SIAM on four benchmark video datasets covering both synthetic and real-world scenarios.

cs.CV↗

arXiv2024

A Shock Flash Breaking Out of a Dusty Red Supergiant

Shock breakout emission is light that arises when a shockwave, generated by core-collapse explosion of a massive star, passes through its outer envelope. Hitherto, the earliest detection of such a signal was at several hours after the explosion, though a few others had been reported. The temporal evolution of early light curves should reveal insights into the shock propagation, including explosion asymmetry and environment in the vicinity, but this has been hampered by the lack of multiwavelength observations. Here we report the instant multiband observations of a type II supernova (SN 2023ixf) in the galaxy M101 (at a distance of 6.85+/-0.15 Mpc), beginning at about 1.4 hours after the explosion. The exploding star was a red supergiant with a radius of about 440 solar radii. The light curves evolved rapidly, on timescales of 1-2 hours, and appeared unusually fainter and redder than predicted by models within the first few hours, which we attribute to an optically thick dust shell before it was disrupted by the shockwave. We infer that the breakout and perhaps the distribution of the surrounding dust were not spherically symmetric.

astro-ph.HE↗

arXiv2024

Online GNN Evaluation Under Test-time Graph Distribution Shifts

Evaluating the performance of a well-trained GNN model on real-world graphs is a pivotal step for reliable GNN online deployment and serving. Due to a lack of test node labels and unknown potential training-test graph data distribution shifts, conventional model evaluation encounters limitations in calculating performance metrics (e.g., test error) and measuring graph data-level discrepancies, particularly when the training graph used for developing GNNs remains unobserved during test time. In this paper, we study a new research problem, online GNN evaluation, which aims to provide valuable insights into the well-trained GNNs's ability to effectively generalize to real-world unlabeled graphs under the test-time graph distribution shifts. Concretely, we develop an effective learning behavior discrepancy score, dubbed LeBeD, to estimate the test-time generalization errors of well-trained GNN models. Through a novel GNN re-training strategy with a parameter-free optimality criterion, the proposed LeBeD comprehensively integrates learning behavior discrepancies from both node prediction and structure reconstruction perspectives. This enables the effective evaluation of the well-trained GNNs' ability to capture test node semantics and structural representations, making it an expressive metric for estimating the generalization error in online GNN evaluation. Extensive experiments on real-world test graphs under diverse graph distribution shifts could verify the effectiveness of the proposed method, revealing its strong correlation with ground-truth test errors on various well-trained GNN models.

cs.LG↗

arXiv2024

Executing Natural Language-Described Algorithms with Large Language Models: An Investigation

Executing computer programs described in natural language has long been a pursuit of computer science. With the advent of enhanced natural language understanding capabilities exhibited by large language models (LLMs), the path toward this goal has been illuminated. In this paper, we seek to examine the capacity of present-day LLMs to comprehend and execute algorithms outlined in natural language. We established an algorithm test set sourced from Introduction to Algorithm, a well-known textbook that contains many representative widely-used algorithms. To systematically assess LLMs' code execution abilities, we selected 30 algorithms, generated 300 random-sampled instances in total, and evaluated whether popular LLMs can understand and execute these algorithms. Our findings reveal that LLMs, notably GPT-4, can effectively execute programs described in natural language, as long as no heavy numeric computation is involved. We believe our findings contribute to evaluating LLMs' code execution abilities and would encourage further investigation and application for the computation power of LLMs.

cs.CL↗

arXiv2024

Graph Learning under Distribution Shifts: A Comprehensive Survey on Domain Adaptation, Out-of-distribution, and Continual Learning

Graph learning plays a pivotal role and has gained significant attention in various application scenarios, from social network analysis to recommendation systems, for its effectiveness in modeling complex data relations represented by graph structural data. In reality, the real-world graph data typically show dynamics over time, with changing node attributes and edge structure, leading to the severe graph data distribution shift issue. This issue is compounded by the diverse and complex nature of distribution shifts, which can significantly impact the performance of graph learning methods in degraded generalization and adaptation capabilities, posing a substantial challenge to their effectiveness. In this survey, we provide a comprehensive review and summary of the latest approaches, strategies, and insights that address distribution shifts within the context of graph learning. Concretely, according to the observability of distributions in the inference stage and the availability of sufficient supervision information in the training stage, we categorize existing graph learning methods into several essential scenarios, including graph domain adaptation learning, graph out-of-distribution learning, and graph continual learning. For each scenario, a detailed taxonomy is proposed, with specific descriptions and discussions of existing progress made in distribution-shifted graph learning. Additionally, we discuss the potential applications and future directions for graph learning under distribution shifts with a systematic analysis of the current state in this field. The survey is positioned to provide general guidance for the development of effective graph learning algorithms in handling graph distribution shifts, and to stimulate future research and advancements in this area.

cs.LG↗

arXiv2024

OVEL: Large Language Model as Memory Manager for Online Video Entity Linking

In recent years, multi-modal entity linking (MEL) has garnered increasing attention in the research community due to its significance in numerous multi-modal applications. Video, as a popular means of information transmission, has become prevalent in people's daily lives. However, most existing MEL methods primarily focus on linking textual and visual mentions or offline videos's mentions to entities in multi-modal knowledge bases, with limited efforts devoted to linking mentions within online video content. In this paper, we propose a task called Online Video Entity Linking OVEL, aiming to establish connections between mentions in online videos and a knowledge base with high accuracy and timeliness. To facilitate the research works of OVEL, we specifically concentrate on live delivery scenarios and construct a live delivery entity linking dataset called LIVE. Besides, we propose an evaluation metric that considers timelessness, robustness, and accuracy. Furthermore, to effectively handle OVEL task, we leverage a memory block managed by a Large Language Model and retrieve entity candidates from the knowledge base to augment LLM performance on memory management. The experimental results prove the effectiveness and efficiency of our method.

cs.CL↗

arXiv2024

Traj-LIO: A Resilient Multi-LiDAR Multi-IMU State Estimator Through Sparse Gaussian Process

Nowadays, sensor suits have been equipped with redundant LiDARs and IMUs to mitigate the risks associated with sensor failure. It is challenging for the previous discrete-time and IMU-driven kinematic systems to incorporate multiple asynchronized sensors, which are susceptible to abnormal IMU data. To address these limitations, we introduce a multi-LiDAR multi-IMU state estimator by taking advantage of Gaussian Process (GP) that predicts a non-parametric continuous-time trajectory to capture sensors' spatial-temporal movement with limited control states. Since the kinematic model driven by three types of linear time-invariant stochastic differential equations are independent of external sensor measurements, our proposed approach is capable of handling different sensor configurations and resilient to sensor failures. Moreover, we replace the conventional $\mathrm{SE}(3)$ state representation with the combination of $\mathrm{SO}(3)$ and vector space, which enables GP-based LiDAR-inertial system to fulfill the real-time requirement. Extensive experiments on the public datasets demonstrate the versatility and resilience of our proposed multi-LiDAR multi-IMU state estimator. To contribute to the community, we will make our source code publicly available.

cs.RO↗

arXiv2023

Mesh-LOAM: Real-time Mesh-Based LiDAR Odometry and Mapping

Despite having achieved real-time performance in mesh construction, most of the current LiDAR odometry and meshing methods may struggle to deal with complex scenes due to relying on explicit meshing schemes. They are usually sensitive to noise. To overcome these limitations, we propose a real-time mesh-based LiDAR odometry and mapping approach for large-scale scenes via implicit reconstruction and a parallel spatial-hashing scheme. To efficiently reconstruct triangular meshes, we suggest an incremental voxel meshing method that updates every scan by traversing each point once and compresses space via a scalable partition module. By taking advantage of rapid accessing triangular meshes at any time, we design point-to-mesh odometry with location and feature-based data association to estimate the poses between the incoming point clouds and the recovered triangular meshes. The experimental results on four datasets demonstrate the effectiveness of our proposed approach in generating accurate motion trajectories and environmental mesh maps.

cs.RO↗