arXiv subjects

Xiao Guo

Publications and source records attributed to Xiao Guo.

At least 19 recordsLinked to original sources

arXiv2026

Gravitational Lensing of Gravitational Wave for Generalized Navarro-Frenk-White Profile and Einasto Profile

The density profiles of Dark matter (DM) halos carry imprints of the DM nature and may be constrained through the lensing effects on gravitational waves (GWs) arising from the halo gravitational potential. In this paper, we investigate GW lensing by two representative types of halo density profiles, i.e., the generalized Navarro-Frenk-White (gNFW) density profile and the Einasto density profile. Using the gravitational lensing equation, we first examine the parameter-space distribution and the imaging characteristics of both profiles under strong lensing, partitioning the parameter space into distinct regions according to the Morse index. We then conduct a detailed analysis of the modulus $\big|F\big|$ and phase $\mathrm{Arg}(F)$ of the amplification factor $F\big(w,y\big)$ at low frequency regime. Our results show that, for a fixed lens mass , increasing the gNFW slope $\gamma$ leads to a larger amplitude, more rapid oscillation, and more distinct wave-packet morphology in the multiple-image regime. Compared with the gNFW case, the Einasto case (with $\alpha=0.16$, $y<0.6$, and the same $M_{200}$) produces a stronger lensing effect. Notably, the evolution of $F\big(w,y\big)$ with frequency for the Einasto profile differs from that of the gNFW case, making its behavior particularly distinctive.

gr-qc↗

arXiv2026

IRGNN: Efficient Invariant Radar Graph Neural Network for Radar Point Cloud Object Detection

Perception is a fundamental component of autonomous driving systems. While LiDAR-based methods have achieved remarkable progress in object detection, their reliability can degrade under adverse weather conditions. Radar point clouds provide a robust alternative due to their resilience to bad weather and low-illumination scenarios. However, radar point clouds are typically sparse, unordered, and less informative than LiDAR data, making it challenging to directly apply existing LiDAR-based perception methods. To address these challenges, we propose IRGNN, an Invariant Radar Graph Neural Network for radar point cloud object detection. IRGNN first reconstructs radar point clouds into graph representations using translation- and rotation-invariant feature designs, enabling robust modeling of sparse radar measurements. It then employs an improved message passing neural network (MPNN) with residual connections and a virtual node layer to enhance local feature propagation and global context modeling. Finally, task-specific heads are applied to the learned graph representations for object classification and bounding box prediction. Experimental results on the RadarScenes dataset show that IRGNN outperforms existing radar-based object detection methods and achieves competitive performance. In addition, IRGNN significantly reduces computational cost and memory usage during inference, demonstrating its effectiveness and practical potential for efficient radar-based perception in autonomous driving.

cs.CV↗

arXiv2026

Multilayer-Dynamic Network Clustering with Application to World Trade Data

The rapid development of global economic integration has made international trade increasingly dynamic and interdependent. The real-world trade data sets, such as the FAO dataset, can be naturally represented as a \emph{multilayer-dynamic network} where countries are treated as nodes, trade flows between countries are represented by edges, and different products correspond to different layers. Therefore, an important problem is how to identify evolving community structures in the multilayer-dynamic trade network. However, most existing methods are designed for static multilayer networks or single-layer dynamic networks, leaving the community detection in multilayer-dynamic networks largely unexplored. Motivated by this problem, we study community detection in multilayer-dynamic networks, allowing the community structure to vary across both layers and time. We propose a novel method, \emph{MuDySC} (Multilayer-Dynamic Spectral Clustering), which smooths the eigenspace projection matrices across adjacent time points and across layers at the same time point. We develop an efficient alternating iterative algorithm for solving the resulting optimization problem and establish its convergence to the global optimum under mild conditions. We further apply MuDySC to the FAO data. The analysis reveals clear asymmetry between export and import community structures and highlights both persistent and shifting trade positions of major countries.

stat.AP↗

arXiv2026

Identifying lensed gravitational waves with physics-informed posterior learning

Gravitational lensing of gravitational waves can probe compact lenses, dark matter substructure, and cosmological distances, but identifying lensed events is difficult when unrelated binary mergers overlap in the same analysis window. We develop physics-informed posterior learning for ranking lensed multi-image signals against unrelated multiple-merger events. The method exploits the geometric-optics consistency that lensing can change amplitudes, arrival times, and Morse phase offsets while preserving the intrinsic phase evolution of the source. We infer a simulation-trained approximate posterior for the common detector-frame chirp mass and symmetric mass ratio, and fuse posterior samples with direct waveform features. Training uses generic multi-image simulations, while point-mass, singular-isothermal-sphere, singular-isothermal-ellipsoid, and shear-perturbed lenses are reserved for held-out lens-family evaluation. For the observationally motivated binary-black-hole population, the fusion ranking raises the detection efficiency from $20.8\%$ to $35.2\%$ at a $1\%$ reference false-positive-rate threshold calibrated on the corresponding unrelated multiple-merger sample. It lowers the network signal-to-noise ratio needed for $50\%$ detection efficiency from 45.3 to 33.5, which corresponds to a 1.35 times larger signal-to-noise-ratio-equivalent distance scale. The gain is limited by loud unrelated multiple-merger events that are partly source consistent, and by the need to calibrate the unrelated multiple-merger population. These results suggest that physical consistency can become a guiding principle for machine learning searches in dense gravitational-wave catalogs.

gr-qc↗

arXiv2026

LC-ERD: Mining Latent Logic for Self-Evolving Reasoning via Consistency-Regulated Reward Decomposition

The evolution of Large Language Model (LLM) reasoning is bottlenecked by the scarcity of high-quality process data. While self-alignment via endogenous rewards offers a solution, mining valid supervision faces three challenges: (1) Label Noise via Mimetic Bias, where rewards prioritize statistical likelihood over logical truth, creating a "correctness illusion" that masks compounding errors; (2) Coarse-Grained Supervision, where sparse global outcomes (e.g., in GRPO) fail to provide granular guidance, treating reasoning chains as monolithic; and (3) Distributional Collapse, where signals fail to generalize without amplifying pre-training biases. To address these, we introduce LC-ERD (Logic-Consistent Endogenous Reward Decomposition), a framework framing self-alignment as latent structure mining. We derive a Variational Logic Potential by aggregating consensus from the model's Latent Logic Expertise (LLE) to denoise the reasoning manifold, and introduce a Multi-Agent Value Decomposition protocol based on the IGM principle to quantify individual step utility. Experiments show LC-ERD delivers a robust self-evolution path, uncovering trade-offs between logic consistency and accuracy while identifying high-value reasoning patterns missed by standard rewards. Our code is available at https://github.com/LC-ERD-repo/LC-ERD.

cs.AI↗

arXiv2026

Transfer Learning for Moderate-Dimensional Ridge-Regularized Robust Linear Regression

This paper studies transfer learning for ridge-regularized robust linear regression in the moderate-dimensional regime, where the number of predictors is of the same order as the sample size and the regression coefficients are not assumed to be sparse. We propose Trans-RR, which combines a robust ridge estimator from a source study with a robust ridge correction based on the target study. Under mild assumptions, we characterize the asymptotic estimation error of the proposed estimator and show that leveraging source data can substantially improve estimation accuracy relative to the traditional single-study ridge-regularized robust estimator. Simulation results and a real-data analysis support the theory and illustrate both positive and negative transfer as the discrepancy between the source and target studies varies.

stat.ME↗

arXiv2026

The Environmental Effects on Inspiraling Binary Black Hole Systems in the Centers of the LMC and M31

Binary black hole (BBH) systems residing in the centers of galaxies evolve within complex astrophysical environments. These environments, comprising dark matter (DM) halos and baryonic accretion disks, can significantly alter the orbital dynamics of the binaries and their resulting gravitational wave (GW) emission. In this study, we investigate the dynamical evolution and GW waveforms of BBH systems embedded in the centers of the Large Magellanic Cloud (LMC) and the Andromeda Galaxy (M31). We construct a comprehensive analytical framework that jointly incorporates GW radiation reaction, DM spike effects (including dynamical friction and accretion, derived from the Navarro-Frenk-White profile), and accretion disk perturbations. Using this framework, we track the long-term evolution of the binary's semi-latus rectum $p$ and orbital eccentricity $e$. Our simulations reveal that the coexistence of a DM spike and an accretion disk significantly accelerates the inspiral process compared to pure DM or vacuum scenarios. Crucially, to assess the observability of these environmental effects, we calculate the Signal-to-Noise Ratio (SNR) and waveform Mismatch for future Pulsar Timing Arrays (PTAs). Our analysis demonstrates that these systems can achieve robust detectability thresholds ($\text{SNR} \ge 8$) within specific parameter spaces. Furthermore, the substantial Mismatch (reaching $\sim 0.7$ over a 20-year observation in the LMC scenario) indicates that the phase deviations induced by these environmental effects are highly distinguishable from vacuum templates. These findings predict the prospect of using future GW detections to probe complex galactic environments.

astro-ph.HE↗

arXiv2026

FusionAgent: A Multimodal Agent with Dynamic Model Selection for Human Recognition

Model fusion is a key strategy for robust recognition in unconstrained scenarios, as different models provide complementary strengths. This is especially important for whole-body human recognition, where biometric cues such as face, gait, and body shape vary across samples and are typically integrated via score-fusion. However, existing score-fusion strategies are usually static, invoking all models for every test sample regardless of sample quality or modality reliability. To overcome these limitations, we propose \textbf{FusionAgent}, a novel agentic framework that leverages a Multimodal Large Language Model (MLLM) to perform dynamic, sample-specific model selection. Each expert model is treated as a tool, and through Reinforcement Fine-Tuning (RFT) with a metric-based reward, the agent learns to adaptively determine the optimal model combination for each test input. To address the model score misalignment and embedding heterogeneity, we introduce Anchor-based Confidence Top-k (ACT) score-fusion, which anchors on the most confident model and integrates complementary predictions in a confidence-aware manner. Extensive experiments on multiple whole-body biometric benchmarks demonstrate that FusionAgent significantly outperforms SoTA methods while achieving higher efficiency through fewer model invocations, underscoring the critical role of dynamic, explainable, and robust model fusion in real-world recognition systems. Project page: \href{https://fusionagent.github.io/}{FusionAgent}.

cs.CV↗

arXiv2026

Detectability of Nearby Binary Neutron Stars with Future sub-mHz Gravitational Wave Missions

Binary neutron stars (BNSs) are one of the most important gravitational wave (GW) sources, which provide key insights to evolution of massive binary stars and nuclear physics. Beyond Laser Interferometer Space Antenna (LISA), Taiji, and Tianqin missions, proposed concepts for next generation space-based GW observatories, including LISAmax, Folkner, and eASTROD, aim to explore the sub-millihertz (mHz) to microhertz ($\mu$ Hz) frequency band. Because the proposed designs substantially suppress low-frequency noise, these detectors are expected to outperform LISA, Taiji, and Tianqin in detecting eccentric Galactic BNS systems. In this paper, we estimate the detectability of nearby inspiraling BNSs using future sub-mHz GW detectors. By utilizing compact binary population synthesis simulations to generate mock BNS samples and estimate their signal-to-noise ratios (SNRs) correspondingly for each GW detector over an observation period of $5-10$\,years, we find that LISAmax may detect $\sim 520-900$ Galactic BNSs, whereas Folkner and eASTROD may detect $\sim 780-1370$ Galactic BNSs. Notably, LISAmax excels in detecting highly eccentric systems $(e>0.90)$ owing to its higher sensitivity at relatively higher sub-mHz frequencies. We further identify seven observed radio BNSs as viable candidates for validation, in particular J0737-3039, which reaches an SNR of $\sim 100$. The expected detection number of LMC inspiraling BNSs is about $\sim 4-18$ for these sub-mHz detectors over an observation period of $5-10$\,years, while detecting inspiraling BNSs in SMC is challenging. This study highlights the significant potential of future sub-mHz GW missions in unraveling BNS formation and evolution physics.

astro-ph.HE↗

arXiv2026

Quantum cascade laser roadmap

Quantum cascade lasers (QCLs) are unipolar semiconductor lasers first demonstrated in 1994. Since then, they have played a central role in advancing mid-infrared and terahertz photonics, becoming among the most reliable light sources in these regions of the electromagnetic spectrum. Their importance is further reinforced by their ability to generate self-starting optical frequency combs, whose investigation is motivated both by fundamental physics and by a wide range of applications, including molecular spectroscopy and free-space optical communications. This Roadmap provides a unified overview of current advances and emerging directions in QCL research. The chapters are organized into three main sections: device design and technology; frequency combs and pulse formation; and applications of QCLs. Each chapter reviews the relevant background, summarizes the current state of the art, and identifies key challenges and future directions within its specific research area.

physics.optics↗

arXiv2026

TRACE: Trajectory-Aware Comprehensive Evaluation for Deep Research Agents

The evaluation of Deep Research Agents is a critical challenge, as conventional outcome-based metrics fail to capture the nuances of their complex reasoning. Current evaluation faces two primary challenges: 1) a reliance on singular metrics like Pass@1, creating a "high-score illusion" that ignores the quality, efficiency, and soundness of the reasoning process; and 2) the failure of static benchmarks to quantify crucial attributes like robustness and latent capability. To address these gaps, we introduce TRACE (Trajectory-Aware Comprehensive Evaluation), a framework that holistically assesses the entire problem-solving trajectory. To counter the "high-score illusion", we propose a Hierarchical Trajectory Utility Function that quantifies process efficiency and cognitive quality, including evidence grounding, alongside accuracy. To measure deeper attributes, TRACE introduces a Scaffolded Capability Assessment protocol, quantifying an agent's latent ability by determining the minimum guidance needed for success. Our contributions include the TRACE framework, its novel metrics, and the accompanying DeepResearch-Bench with controllable complexity. Experiments show TRACE delivers a granular ranking that uncovers critical trade-offs between agent accuracy, efficiency, and robustness entirely missed by singular metrics.

cs.CL↗

arXiv2026

On the Holistic Approach for Detecting Human Image Forgery

The rapid advancement of AI-generated content (AIGC) has escalated the threat of deepfakes, from facial manipulations to the synthesis of entire photorealistic human bodies. However, existing detection methods remain fragmented, specializing either in facial-region forgeries or full-body synthetic images, and consequently fail to generalize across the full spectrum of human image manipulations. We introduce HuForDet, a holistic framework for human image forgery detection, which features a dual-branch architecture comprising: (1) a face forgery detection branch that employs heterogeneous experts operating in both RGB and frequency domains, including an adaptive Laplacian-of-Gaussian (LoG) module designed to capture artifacts ranging from fine-grained blending boundaries to coarse-scale texture irregularities; and (2) a contextualized forgery detection branch that leverages a Multi-Modal Large Language Model (MLLM) to analyze full-body semantic consistency, enhanced with a confidence estimation mechanism that dynamically weights its contribution during feature fusion. We curate a human image forgery (HuFor) dataset that unifies existing face forgery data with a new corpus of full-body synthetic humans. Extensive experiments show that our HuForDet achieves state-of-the-art forgery detection performance and superior robustness across diverse human image forgeries.

cs.CV↗

arXiv2025

Multiband Gravitational Wave Detection Prospects for M31 UCXB-1 System in Low and Middle Frequency Band

The recent discovery of M31 UCXB-1, the first extragalactic ultracompact X-ray binary (UCXB) with an orbital period of $T_{\rm orb} \sim 465$ s, presents a unique laboratory for studying close binary evolution and an unprecedented target for continuous gravitational wave (GW) searches. Its identification as a strong candidate black hole-white dwarf (BH-WD) system, combined with its exceptionally short period and high X-ray luminosity, suggests it may be one of the most vital low-frequency GW sources in M31. In this \textit{Letter}, we investigate the detectability of its GW signal for future space-borne detectors in multiband GW detection. We find that while its signal-to-noise ratio (S/N) for low-frequency detectors remains marginal for high-confidence detection, middle-frequency detectors such as DECIGO and BBO are far more promising, potentially achieving S/N $\varrho>8$ within reasonable observational duration. With a primary mass of only $m_1 > 5.4M_\odot$ (or $6.6M_\odot$), the network of all low and middle frequency detector (or BBO alone) is capable of detecting GW from this system with a $\varrho > 8$, during 10-year observation. Furthermore, orbital eccentricity can enhance the GW strain at higher harmonics, further improving its detectability, especially for middle-frequency detectors. This study establishes M31 UCXB-1 as a key prototype of short-period UCXBs, cementing its role as a cornerstone for multiband, multi-messenger astrophysics and a vital bridge between X-ray astronomy and the future GW era.

astro-ph.HE↗

arXiv2025

Estimation of gravitational wave from solar emerging magnetic flux tube

This study investigates the gravitational waves (GWs) generated by the emergence of magnetic flux tubes in the solar convection zone. We focus on the upward buoyancy of magnetic flux tubes, which leads to significant magnetic activity and the formation of active region sunspots. This study adopts parameters representative of a moderate-sized solar active region to estimate the GWs generated by the emergence of magnetic flux tubes. Our results indicate that the GW strain amplitude, achievable through signal superposition and detection at close proximity (e.g., approximately one solar radius from the solar surface), may reach $\sim$10$^{-29}$. The characteristic GW frequency is estimated at $\sim$10$^{-5}$ Hz, placing it at the high-frequency end of the sensitivity band of Pulsar Timing Array (PTA) methods. However, the estimated strain amplitudes remain orders of magnitude below the sensitivity thresholds of current and foreseeable gravitational wave detectors. Notably, reducing the cadence $\Delta t$ of Pulsar Timing Array (PTA) observations to approximately 2 hours ($\Delta t = 2\text{hours}$) would raise the maximum detectable frequency to about $5.8 \times 10^{-5} \text{Hz}$, thereby encompassing the dominant spectral component of solar activity-related GWs predicted in this study, offering a potential pathway for future detection. Successful detection in the future may help to predict the super solar active region emergence in space weather forecasting.

astro-ph.SR↗

arXiv2025

Enhanced Localization of Dark Lensed Gravitational Wave Events Enables Host Galaxy Identification and Precise Cosmological Inference

Lensed gravitational wave (GW) events are expected to be powerful new probes of cosmology, contingent on redshift measurement by electromagnetic observations. Host galaxy identification is thus crucial but challenging due to poor localization by GW signal alone. In this paper, we show that the third-generation ground-based GW detectors will detect a population of lensed events with three or more detectable images (including the central one), each arriving at distinct times and Earth locations in the space, forming an effective network that reduces the typical localization area to $\sim0.01$ deg$^2$. For at least $90\%$ (or $50\%$) of these events, the localization improves by more than a factor of $10$ (or $30$) comparing with unlensed cases. Such precise localization and multiple-image detections enable robust host-galaxy identification and, through lens modelling, further yield sub-arcsecond position. As ``dark lensed sirens", these events become powerful probes of cosmological parameters. Using simulated lensed compact-binary mergers, we show that two-year or longer observations with third-generation GW detectors can measure the Hubble constant to $\lesssim1$\% precision via ``dark lensed sirens" (even when relying solely on lensed stellar-mass binary black hole events), while simultaneously constraining other cosmological parameters. This approach will provide an independent, complementary avenue for measuring cosmological parameters.

astro-ph.CO↗

arXiv2025

Multi-Channel Amplitude-Phase Asymmetric-Encrypted Janus Acoustic Meta-Holograms

Encrypted optical and acoustic meta-holograms only focus on the encrypted hologram in a single channel, viz. modulating spatial amplitude to project a holographic image. In this research, the unique concept of multi-channel amplitude-phase asymmetric-encrypted Janus acoustic meta-holograms is proposed, demonstrating remarkable capabilities of generating, encrypting, and decrypting both amplitude and phase holographic images on both sides of a metascreen. The flexible and decoupled manipulation mechanism for the amplitude-phase of the bidirectional acoustic waves used in our concept offers multiple possibilities to apply various encryption methods. In this work, our system enables single-input, two-faced four-channel asymmetric encryption, which substantially increase the communication capacity of conventional acoustic holograms, and establish a security framework based on mathematical problem, proving its security. Our work can lead to concrete applications including, but not limited to, multi-channel acoustic field communications and acoustic illusion and cloaking in non-transparent media.

physics.app-ph↗

arXiv2025

Bright siren without electromagnetic counterpart by LISA-Taiji-TianQin network

Gravitational waves (GWs) with electromagnetic counterparts (EMc) offer a novel approach to measure the Hubble constant ($H_0$), known as bright sirens, enabling $H_0$ measurements by combining GW-derived distances with EM-derived redshifts. Host galaxy identification is essential for redshift determination but remains challenging due to poor GW sky localization and uncertainties in EMc models. To overcome these limitations, we exploit the ultra-high-precision localization ($\Delta \Omega_s \sim 10^{-4} \, \text{deg}^2$) with a space-based GW detector network (LISA-Taiji-TianQin), which permits unique host identification solely from GW signals. We integrate five massive black hole binary (MBHB) population models and two galaxy number density models to compute the redshift horizon for host galaxy identification and evaluate $H_0$ constraints. We find that (1) The network enhances localization by several orders of magnitude compared to single detectors; (2) The identification horizon reaches $z\sim 1.2$ for specific MBHBs in the most accurate localization case; (3) The population model choice critically impacts the outcomes: the most refined population models yield to independent EMc identification rate of 0.6-1 $\text{yr}^{-1}$ with $H_0$ constraints $< 1\%$ fractional uncertainty, the less refined models lead to the rate $<0.1\text{yr}^{-1}$ and $1-2\%$ uncertainty on $H_0$.

astro-ph.CO↗

arXiv2025

Prospective constraints on dark energy from nanohertz individual gravitational wave sources

Nanohertz gravitational waves (GWs) from supermassive binary black holes (SMBBHs), detectable via pulsar timing arrays (PTAs), offer a novel avenue to constrain dark energy. Based on cosmological simulations and semi-analytic galaxy formation models, this study explores the detectability of individual nanohertz SMBBH sources using next-generation PTAs and their potential for constraining dark energy under an optimistic scenario considering only the presence of white noise. By constructing light-cone SMBBH populations across hardening timescales ($\tau_H = 0.1/5/10$Gyr) and computing signal-to-noise ratios (SNR), we find advanced PTAs can resolve $10^2$--$10^3$ sources with SNR $> 8$ (primarily at $z < 1$ with chirp masses of $10^8$--$10^{10}M_{\odot}$). If electromagnetic counterparts can be identified, optimal configurations ($\sigma_t = 50$ns, $N_p = 1000$, $T_{\text{obs}} = 30$yr with$ \tau_H \leq 5$Gyr) could constrain the dark energy equation-of-state (EoS) parameter $w$ to $\Delta w \sim 0.023$--$0.048$, where the constraints only exhibit weak dependence on $\tau_H$ within $0.1$--$5$Gyr. If only $10\%$ of GW sources have detectable electromagnetic counterparts, constraints weaken to $\Delta w = 0.075$ ($\tau_H = 0.1$Gyr) and $\Delta w = 0.162$ ($\tau_H = 5$Gyr) under the most optimal parameter configuration. What's more, conservative PTAs ($N_p = 500$, $\sigma_t = 100$--$200$ns) with additional $30$-year data accumulation could double resolvable source counts and improve $\Delta w$ precision by $\sim 40\%$.

astro-ph.CO↗