arXiv subjects

Zhan Gao

Publications and source records attributed to Zhan Gao.

At least 19 recordsLinked to original sources

arXiv2026

Probing Saturation Effect in Heavy Meson Pair Correlation in Forward $pA$ Collisions

Forward two-particle angular correlations in $pA$ collisions have long been recognized as a particularly sensitive observable for exploring gluon saturation effects. In the back-to-back regime, two-particle correlations receive substantial contributions from both soft-gluon radiation and saturation effects. In this work, we study heavy meson pair correlation in forward proton-nucleus collisions by incorporating a unified Sudakov resummation for heavy meson pair correlations in the Color Glass Condensate effect theory. Our results are in good agreement with the $\Delta\phi$ data measured by the LHCb Collaboration for $D^0 \bar D^0$ pairs in forward $pp$ and $pA$ collisions, as well as $J/\psi$ pairs from $b\bar b$ decays in forward $pp$ collisions. Furthermore, we present predictions for $D\bar D$ and $B\bar B$ correlations in the forward rapidity regions at the Large Hadron Collider. A pronounced mass-hierarchy is observed in the nuclear modification factor, $R_{pA}\big|_{m_b}<R_{pA}\big|_{m_c}$, indicating stronger sensitivity to saturation effects at small $x$. As the rapidity increases, the suppression becomes more pronounced while the mass hierarchy remains robust. This study will help us to search for the saturation signal via heavy-meson pair correlations in forward $pA$ collisions.

hep-ph↗

arXiv2026

Differentiable Environment-Trajectory Co-Optimization for Safe Multi-Agent Navigation

The environment plays a critical role in multi-agent navigation by imposing spatial constraints, rules, and limitations that agents must navigate around. Traditional approaches treat the environment as fixed, without exploring its impact on agents' performance. This work considers environment configurations as decision variables, alongside agent actions, to jointly achieve safe navigation. We formulate a bi-level problem, where the lower-level sub-problem optimizes agent trajectories that minimize navigation cost and the upper-level sub-problem optimizes environment configurations that maximize navigation safety. We develop a differentiable optimization method that iteratively solves the lower-level sub-problem with interior point methods and the upper-level sub-problem with gradient ascent. A key challenge lies in analytically coupling these two levels. We address this by leveraging KKT conditions and the Implicit Function Theorem to compute gradients of agent trajectories w.r.t. environment parameters, enabling differentiation throughout the bi-level structure. Moreover, we propose a novel metric that quantifies navigation safety as a criterion for the upper-level environment optimization, and prove its validity through measure theory. Our experiments validate the effectiveness of the proposed framework in a variety of safety-critical navigation scenarios, inspired from warehouse logistics to urban transportation. The results demonstrate that optimized environments provide navigation guidance, improving both agents' safety and efficiency.

cs.RO↗

arXiv2026

Stochastic Sequential Decision Making over Expanding Networks with Graph Filtering

Graph filters leverage topological information to process networked data with existing methods mainly studying fixed graphs, ignoring that graphs often expand as nodes continually attach with an unknown pattern. The latter requires developing filter-based decision-making paradigms that take evolution and uncertainty into account. Existing approaches rely on either pre-designed filters or online learning, limited to a myopic view considering only past or present information. To account for future impacts, we propose a stochastic sequential decision-making framework for filtering networked data with a policy that adapts filtering to expanding graphs. By representing filter shifts as agents, we model the filter as a multi-agent system and train the policy following multi-agent reinforcement learning. This accounts for long-term rewards and captures expansion dynamics through sequential decision-making. Moreover, we develop a context-aware graph neural network to parameterize the policy, which tunes filter parameters based on information of both the graph and agents. Experiments on synthetic and real datasets from cold-start recommendation to COVID prediction highlight the benefits of using a sequential decision-making perspective over batch and online filtering alternatives.

cs.LG↗

arXiv2026

Goal-Oriented Learning at the Edge: Graph Neural Networks Over-the-Air for Blockage Prediction

Sixth-generation (6G) wireless networks evolve from connecting devices to connecting intelligence. The focus turns to Goal-Oriented Communications, where the effectiveness of communication is assessed through task-level objectives over traditional throughput-centric metrics. As communication intertwines with learning at the edge, distributed inference over wireless networks faces a critical trade-off between task accuracy and efficient radio resource use. Traditional communication schemes (e.g., OFDMA) are not designed for this trade-off, often facing challenges related to scalability and latency. Therefore, we propose a novel goal-oriented framework that integrates over-the-air computation with spatio-temporal graph learning. Leveraging the wireless channel as an analog aggregation layer, the proposed framework enables low-latency message passing while efficiently aggregating semantically relevant features from distributed nodes. Theoretical analysis confirms that our analog architecture converges to the expressive power of digital message passing, while offering decisive scalability advantages. We assess the framework in proactive line-of-sight blockage prediction for millimeter-wave networks. Through high-fidelity ray-tracing simulations, the framework exhibits strong inductive generalization to unseen networks and adapts to domain shifts via lightweight transfer learning, matching or even outperforming digital baselines with significantly reduced communication overhead.

cs.NI↗

arXiv2026

MPA: Multimodal Prototype Augmentation for Few-Shot Learning

Recently, few-shot learning (FSL) has become a popular task that aims to recognize new classes from only a few labeled examples and has been widely applied in fields such as natural science, remote sensing, and medical images. However, most existing methods focus only on the visual modality and compute prototypes directly from raw support images, which lack comprehensive and rich multimodal information. To address these limitations, we propose a novel Multimodal Prototype Augmentation FSL framework called MPA, including LLM-based Multi-Variant Semantic Enhancement (LMSE), Hierarchical Multi-View Augmentation (HMA), and an Adaptive Uncertain Class Absorber (AUCA). LMSE leverages large language models to generate diverse paraphrased category descriptions, enriching the support set with additional semantic cues. HMA exploits both natural and multi-view augmentations to enhance feature diversity (e.g., changes in viewing distance, camera angles, and lighting conditions). AUCA models uncertainty by introducing uncertain classes via interpolation and Gaussian sampling, effectively absorbing uncertain samples. Extensive experiments on four single-domain and six cross-domain FSL benchmarks demonstrate that MPA achieves superior performance compared to existing state-of-the-art methods across most settings. Notably, MPA surpasses the second-best method by 12.29% and 24.56% in the single-domain and cross-domain setting, respectively, in the 5-way 1-shot setting.

cs.CV↗

arXiv2025

Hierarchical Dual-Strategy Unlearning for Biomedical and Healthcare Intelligence Using Imperfect and Privacy-Sensitive Medical Data

Large language models (LLMs) exhibit exceptional performance but pose substantial privacy risks due to training data memorization, particularly within healthcare contexts involving imperfect or privacy-sensitive patient information. We present a hierarchical dual-strategy framework for selective knowledge unlearning that precisely removes specialized knowledge while preserving fundamental medical competencies. Our approach synergistically integrates geometric-constrained gradient updates to selectively modulate target parameters with concept-aware token-level interventions that distinguish between preservation-critical and unlearning-targeted tokens via a unified four-level medical concept hierarchy. Comprehensive evaluations on the MedMCQA (surgical) and MHQA (anxiety, depression, trauma) datasets demonstrate superior performance, achieving an 82.7% forgetting rate and 88.5% knowledge preservation. Notably, our framework maintains robust privacy guarantees while requiring modification of only 0.1% of parameters, addressing critical needs for regulatory compliance, auditability, and ethical standards in clinical research.

cs.LG↗

arXiv2025

Learning a Decentralized Medium Access Control Protocol for Shared Message Transmission

In large-scale Internet of things networks, efficient medium access control (MAC) is critical due to the growing number of devices competing for limited communication resources. In this work, we consider a new challenge in which a set of nodes must transmit a set of shared messages to a central controller, without inter-node communication or retransmissions. To our knowledge, this is the first work to formalize and analyze the multi-shared-message decentralized random access problem. Messages are distributed among random subsets of nodes, which must implicitly coordinate their transmissions over shared communication opportunities. The objective is to guarantee the delivery of all shared messages, regardless of which nodes transmit them. We first prove the existence of an optimal deterministic strategy, and characterize the success rate degradation of a deterministic strategy under dynamic message-transmission patterns. To solve this problem, we propose a decentralized learning-based framework that enables nodes to autonomously synthesize deterministic transmission strategies aiming to maximize message delivery success, together with an online adaptation mechanism that maintains stable performance in dynamic scenarios. Extensive simulations validate the framework's effectiveness, scalability, and adaptability, demonstrating its robustness to varying network sizes and fast adaptation to dynamic changes in transmission patterns, outperforming state-of-the-art approaches.

cs.NI↗

arXiv2025

ReCoDe: Reinforcement Learning-based Dynamic Constraint Design for Multi-Agent Coordination

Constraint-based optimization is a cornerstone of robotics, enabling the design of controllers that reliably encode task and safety requirements such as collision avoidance or formation adherence. However, handcrafted constraints can fail in multi-agent settings that demand complex coordination. We introduce ReCoDe--Reinforcement-based Constraint Design--a decentralized, hybrid framework that merges the reliability of optimization-based controllers with the adaptability of multi-agent reinforcement learning. Rather than discarding expert controllers, ReCoDe improves them by learning additional, dynamic constraints that capture subtler behaviors, for example, by constraining agent movements to prevent congestion in cluttered scenarios. Through local communication, agents collectively constrain their allowed actions to coordinate more effectively under changing conditions. In this work, we focus on applications of ReCoDe to multi-agent navigation tasks requiring intricate, context-based movements and consensus, where we show that it outperforms purely handcrafted controllers, other hybrid approaches, and standard MARL baselines. We give empirical (real robot) and theoretical evidence that retaining a user-defined controller, even when it is imperfect, is more efficient than learning from scratch, especially because ReCoDe can dynamically change the degree to which it relies on this controller.

cs.RO↗

arXiv2025

Extracting Essential Non-perturbative Information in Jet Invariant Mass via the Bayesian Analysis

In this paper, we present a new three-dimensional non-perturbative (NP) function to account for and parameterize the NP contributions in the jet invariant mass spectrum, in addition to the conventional NP mass shift parametrization. By implementing Bayesian analysis on experimental data of the jet invariant mass from exclusive $W/Z+$jet events and inclusive jet events in $pp$ collisions at RHIC and LHC, where collisional energy increases by a factor of up to $65$ from RHIC to LHC, we ensure the analysis covers a wide range of data. For the first time, we simultaneously extract NP contributions from hadronization, initial soft-gluon radiation, and underlying events, based on two different NP prescriptions. We find that the contribution from initial soft-gluon radiation is negligible, and the hadronization effect dominates in the small-$R$ region, while underlying events provide the dominant contribution in the large-$R$ region. Moreover, when only hadronization effects are considered, our results successfully describe the jet mass data measured in $e^+e^-$ collisions, where only hadronization effects are expected to be present. Our work offers quantitative insights into understanding the soft hadronic contribution to jet substructure.

hep-ph↗

arXiv2025

Bayesian inference of the magnetic field and chemical potential on holographic jet quenching in heavy-ion collisions

Jet quenching is studied in a background magnetic field and a finite baryon chemical potential. The production of energetic partons is calculated using the next-to-leading order (NLO) perturbative Quantum Chromodynamics (pQCD) parton model, while the parton energy loss formula is obtained from the AdS/CFT correspondence incorporating the magnetic field and baryon chemical potential effects. Using Bayesian inference, we systemically compare the theoretical calculations with experimental data for the nuclear modification factor $R_{AA}$ of the large transverse momentum hadrons in different-centrality nucleus-nucleus collisions at 0.2, 2.76 and 5.02 TeV, respectively. The form of the holographic energy loss leads to a strong negative correlation between the magnetic field and the chemical potential after the calibration, from which we discussed the sensitivity of jet quenching phenomena to magnetic field and baryon chemical potential.

hep-ph↗

arXiv2025

Tianyi: A Traditional Chinese Medicine all-rounder language model and its Real-World Clinical Practice

Natural medicines, particularly Traditional Chinese Medicine (TCM), are gaining global recognition for their therapeutic potential in addressing human symptoms and diseases. TCM, with its systematic theories and extensive practical experience, provides abundant resources for healthcare. However, the effective application of TCM requires precise syndrome diagnosis, determination of treatment principles, and prescription formulation, which demand decades of clinical expertise. Despite advancements in TCM-based decision systems, machine learning, and deep learning research, limitations in data and single-objective constraints hinder their practical application. In recent years, large language models (LLMs) have demonstrated potential in complex tasks, but lack specialization in TCM and face significant challenges, such as too big model scale to deploy and issues with hallucination. To address these challenges, we introduce Tianyi with 7.6-billion-parameter LLM, a model scale proper and specifically designed for TCM, pre-trained and fine-tuned on diverse TCM corpora, including classical texts, expert treatises, clinical records, and knowledge graphs. Tianyi is designed to assimilate interconnected and systematic TCM knowledge through a progressive learning manner. Additionally, we establish TCMEval, a comprehensive evaluation benchmark, to assess LLMs in TCM examinations, clinical tasks, domain-specific question-answering, and real-world trials. The extensive evaluations demonstrate the significant potential of Tianyi as an AI assistant in TCM clinical practice and research, bridging the gap between TCM knowledge and practical application.

cs.CL↗

arXiv2024

Sparse Covariance Neural Networks

Covariance Neural Networks (VNNs) perform graph convolutions on the covariance matrix of input data to leverage correlation information as pairwise connections. They have achieved success in a multitude of applications such as neuroscience, financial forecasting, and sensor networks. However, the empirical covariance matrix on which VNNs operate typically contains spurious correlations, creating a mismatch with the actual covariance matrix that degrades VNNs' performance and computational efficiency. To tackle this issue, we put forth Sparse coVariance Neural Networks (S-VNNs), a framework that applies sparsification techniques on the sample covariance matrix and incorporates the latter into the VNN architecture. We investigate the S-VNN when the underlying data covariance matrix is both sparse and dense. When the true covariance matrix is sparse, we propose hard and soft thresholding to improve the covariance estimation and reduce the computational cost. Instead, when the true covariance is dense, we propose a stochastic sparsification where data correlations are dropped in probability according to principled strategies. Besides performance and computation improvements, we show that S-VNNs are more stable to finite-sample covariance estimations than nominal VNNs and the analogous sparse principal component analysis. By analyzing the impact of sparsification on their behavior, we tie the S-VNN stability to the data distribution and sparsification approach. We support our theoretical findings with experimental results on a variety of application scenarios, ranging from brain data to human action recognition, and show an improved task performance, improved stability, and reduced computational time compared to alternatives.

cs.LG↗

arXiv2024

LASSO Inference for High Dimensional Predictive Regressions

LASSO inflicts shrinkage bias on estimated coefficients, which undermines asymptotic normality and invalidates standard inferential procedures based on the t-statistic. Given cross sectional data, the desparsified LASSO has emerged as a well-known remedy for correcting the shrinkage bias. In the context of high dimensional predictive regression, the desparsified LASSO faces an additional challenge: the Stambaugh bias arising from nonstationary regressors modeled as local unit roots. To restore standard inference, we propose a novel estimator called IVX-desparsified LASSO (XDlasso). XDlasso simultaneously eliminates both shrinkage bias and Stambaugh bias and does not require prior knowledge about the identities of nonstationary and stationary regressors. We establish the asymptotic properties of XDlasso for hypothesis testing, and our theoretical findings are supported by Monte Carlo simulations. Applying our method to real-world applications from the FRED-MD database, we investigate two important empirical questions: (i) the predictability of the U.S. stock returns based on the earnings-price ratio, and (ii) the predictability of the U.S. inflation using the unemployment.

stat.ME↗

arXiv2024

Robust Estimation of Regression Models with Potentially Endogenous Outliers via a Modern Optimization Lens

This paper addresses the robust estimation of linear regression models in the presence of potentially endogenous outliers. Through Monte Carlo simulations, we demonstrate that existing $L_1$-regularized estimation methods, including the Huber estimator and the least absolute deviation (LAD) estimator, exhibit significant bias when outliers are endogenous. Motivated by this finding, we investigate $L_0$-regularized estimation methods. We propose systematic heuristic algorithms, notably an iterative hard-thresholding algorithm and a local combinatorial search refinement, to solve the combinatorial optimization problem of the $L_0$-regularized estimation efficiently. Our Monte Carlo simulations yield two key results: (i) The local combinatorial search algorithm substantially improves solution quality compared to the initial projection-based hard-thresholding algorithm while offering greater computational efficiency than directly solving the mixed integer optimization problem. (ii) The $L_0$-regularized estimator demonstrates superior performance in terms of bias reduction, estimation accuracy, and out-of-sample prediction errors compared to $L_1$-regularized alternatives. We illustrate the practical value of our method through an empirical application to stock return forecasting.

econ.EM↗

arXiv2024

Graph Neural Networks over the Air for Decentralized Tasks in Wireless Networks

Graph neural networks (GNNs) model representations from networked data and allow for decentralized inference through localized communications. Existing GNN architectures often assume ideal communications and ignore potential channel effects, such as fading and noise, leading to performance degradation in real-world implementation. Considering a GNN implemented over nodes connected through wireless links, this paper conducts a stability analysis to study the impact of channel impairments on the performance of GNNs, and proposes graph neural networks over the air (AirGNNs), a novel GNN architecture that incorporates the communication model. AirGNNs modify graph convolutional operations that shift graph signals over random communication graphs to take into account channel fading and noise when aggregating features from neighbors, thus, improving architecture robustness to channel impairments during testing. We develop a channel-inversion signal transmission strategy for AirGNNs when channel state information (CSI) is available, and propose a stochastic gradient descent based method to train AirGNNs when CSI is unknown. The convergence analysis shows that the training procedure approaches a stationary solution of an associated stochastic optimization problem and the variance analysis characterizes the statistical behavior of the trained model. Experiments on decentralized source localization and multi-robot flocking corroborate theoretical findings and show superior performance of AirGNNs over wireless communication channels.

eess.SP↗

arXiv2024

Co-Optimizing Reconfigurable Environments and Policies for Decentralized Multi-Agent Navigation

This work views the multi-agent system and its surrounding environment as a co-evolving system, where the behavior of one affects the other. The goal is to take both agent actions and environment configurations as decision variables, and optimize these two components in a coordinated manner to improve some measure of interest. Towards this end, we consider the problem of decentralized multi-agent navigation in a cluttered environment, where we assume that the layout of the environment is reconfigurable. By introducing two sub-objectives -- multi-agent navigation and environment optimization -- we propose an agent-environment co-optimization problem and develop a coordinated algorithm that alternates between these sub-objectives to search for an optimal synthesis of agent actions and environment configurations; ultimately, improving the navigation performance. Due to the challenge of explicitly modeling the relation between the agents, the environment and their performance therein, we leverage policy gradient to formulate a model-free learning mechanism within the coordinated framework. A formal convergence analysis shows that our coordinated algorithm tracks the local minimum solution of an associated time-varying non-convex optimization problem. Experiments corroborate theoretical findings and show the benefits of co-optimization. Interestingly, the results also indicate that optimized environments can offer structural guidance to de-conflict agents in motion.

cs.RO↗

arXiv2023

Momentum and angular correlations in \texorpdfstring{$Z/γ$}{Z/gamma}-hadron production in relativistic heavy-ion collisions

We carry out a detailed study of medium modifications on momentum and angular correlations between a large transverse momentum hadron and a $Z/γ$ trigger in relativistic heavy-ion collisions within a perturbative QCD parton model improved by the Sudakov resummation technique. The total energy loss of a hard parton propagating inside the medium is employed to modify the fragmentation function, while the medium-induced transverse momentum broadening is included in the resummation approach, and both of them are related to the jet transport parameter and obtained by the high-twist formalism. We obtain good agreements with the existing data on transverse momentum and azimuthal angular correlations for the $Z/γ$-hadron pairs in $pp$ and $AA$ collisions, and predict the correlations for the $γ$-hadron in central $PbPb$ collisions at 5.02 TeV. The numerical analyses for the $Z/γ$-hadron in central $PbPb$ collisions show that the normalized angular distribution is decorrelated due to the medium-induced transverse momentum broadening, however, the angular correlation is enhanced due to the parton energy loss, namely anti-broadening. The observed modification of the angular correlation is a result of the competition between the broadening and the anti-broadening. This work provides a reliable theoretical tool for a comprehensive and precise study of jet quenching in relativistic heavy-ion collisions.

hep-ph↗

arXiv2023

On the Trade-Off between Stability and Representational Capacity in Graph Neural Networks

Analyzing the stability of graph neural networks (GNNs) under topological perturbations is key to understanding their transferability and the role of each architecture component. However, stability has been investigated only for particular architectures, questioning whether it holds for a broader spectrum of GNNs or only for a few instances. To answer this question, we study the stability of EdgeNet: a general GNN framework that unifies more than twenty solutions including the convolutional and attention-based classes, as well as graph isomorphism networks and hybrid architectures. We prove that all GNNs within the EdgeNet framework are stable to topological perturbations. By studying the effect of different EdgeNet categories on the stability, we show that GNNs with fewer degrees of freedom in their parameter space, linked to a lower representational capacity, are more stable. The key factor yielding this trade-off is the eigenvector misalignment between the EdgeNet parameter matrices and the graph shift operator. For example, graph convolutional neural networks that assign a single scalar per signal shift (hence, with a perfect alignment) are more stable than the more involved node or edge-varying counterparts. Extensive numerical results corroborate our theoretical findings and highlight the role of different architecture components in the trade-off.

eess.SP↗