arXiv subjects

Xin Wang

Publications and source records attributed to Xin Wang.

At least 523 records · Page 29Linked to original sources

arXiv2025

Quantifying Spectroscopic Flux Variations Between JWST NIRISS and NIRSpec: Slit Losses in Emission Line Measurements of z$\sim$1-3 Galaxies

We analyze JWST NIRISS and NIRSpec spectroscopic observations in the Abell 2744 galaxy cluster field. From approximately 120 candidates, we identify 12 objects with at least a prominent emission lines among \Oii, \Hb, \Oiiia, \Oiiib, and \Ha that are spectroscopically confirmed by both instruments. Our key findings reveal systematic differences between the two spectrographs based on source morphology and shutter aperture placement. Compact objects show comparable or higher integrated flux in NIRSpec relative to NIRISS (within 1$σ$ uncertainties), while extended sources consistently display higher flux in NIRISS measurements. This pattern reflects NIRSpec's optimal coverage for compact objects while potentially undersampling extended sources. Quantitative analysis demonstrates that NIRSpec recovers at least $63\%$ of NIRISS-measured flux when the slit covers $>15\%$ of the source or when $R_e<1$kpc. For lower coverage or larger effective radii, the recovered flux varies from $24\%$ to $63\%$. When studying the \Ha/\Oiiib emission line ratio, we observe that measurements from these different spectrographs can vary by up to $\sim$0.3 dex, with significant implications for metallicity and star formation rate characterizations for individual galaxies. These results highlight the importance of considering instrumental effects when combining multi-instrument spectroscopic data and demonstrate that source morphology critically influences flux recovery between slit-based and slitless spectroscopic modes in JWST observations.

astro-ph.GA↗

arXiv2025

Tight Generalization Bound for Supervised Quantum Machine Learning

We derive a tight generalization bound for quantum machine learning that is applicable to a wide range of supervised tasks, data, and models. Our bound is both efficiently computable and free of big-O notation. Furthermore, we point out that previous bounds relying on big-O notation may provide misleading suggestions regarding the generalization error. Our generalization bound demonstrates that for quantum machine learning models of arbitrary size and depth, the sample size is the most dominant factor governing the generalization error. Additionally, the spectral norm of the measurement observable, the bound and Lipschitz constant of the selected risk function also influence the generalization upper bound. However, the number of quantum gates, the number of qubits, data encoding methods, and hyperparameters chosen during the learning process such as batch size, epochs, learning rate, and optimizer do not significantly impact the generalization capability of quantum machine learning. We experimentally demonstrate the tightness of our generalization bound across classification and regression tasks. Furthermore, we show that our tight generalization upper bound holds even when labels are completely randomized. We thus bring clarity to the fundamental question of generalization in quantum machine learning.

quant-ph↗

arXiv2025

Cosmological Consequences of Domain Walls Biased by Quantum Gravity

One of the simplest standard model extensions leading to a domain wall network is a real scalar $S$ with a $Z_2$ symmetry spontaneously broken during universe evolution. Motivated by the swampland program, we explore the possibility that quantum gravity effects are responsible for violation of the discrete symmetry, triggering the annihilation of the domain wall network. We explore the resulting cosmological implications in terms of dark radiation, dark matter, gravitational waves, primordial black holes, and wormholes connected to baby universes.

hep-ph↗

arXiv2025

Revisiting Transformation Invariant Geometric Deep Learning: An Initial Representation Perspective

Deep neural networks have achieved great success in the last decade. When designing neural networks to handle the ubiquitous geometric data such as point clouds and graphs, it is critical that the model can maintain invariance towards various transformations such as translation, rotation, and scaling. Most existing graph neural network (GNN) approaches can only maintain permutation-invariance, failing to guarantee invariance with respect to other transformations. Besides GNNs, other works design sophisticated transformation-invariant layers, which are computationally expensive and difficult to be extended. In this paper, we revisit why general neural networks cannot maintain transformation invariance. Our findings show that transformation-invariant and distance-preserving initial point representations are sufficient to achieve transformation invariance rather than needing sophisticated neural layer designs. Motivated by these findings, we propose Transformation Invariant Neural Networks (TinvNN), a straightforward and general plug-in for geometric data. Specifically, we realize transformation invariant and distance-preserving initial point representations by modifying multi-dimensional scaling and feed the representations into existing neural networks. We prove that TinvNN can strictly guarantee transformation invariance, being general and flexible enough to be combined with the existing neural networks. Extensive experimental results on point cloud analysis and combinatorial optimization demonstrate the effectiveness and general applicability of our method. We also extend our method into equivariance cases. Based on the results, we advocate that TinvNN should be considered as an essential baseline for further studies of transformation-invariant geometric deep learning.

cs.CV↗

arXiv2025

Image-Plane Geometric Decoding for View-Invariant Indoor Scene Reconstruction

Volume-based indoor scene reconstruction methods offer superior generalization capability and real-time deployment potential. However, existing methods rely on multi-view pixel back-projection ray intersections as weak geometric constraints to determine spatial positions. This dependence results in reconstruction quality being heavily influenced by input view density. Performance degrades in overlapping regions and unobserved areas.To address these limitations, we reduce dependency on inter-view geometric constraints by exploiting spatial information within individual views. We propose an image-plane decoding framework with three core components: Pixel-level Confidence Encoder, Affine Compensation Module, and Image-Plane Spatial Decoder. These modules decode three-dimensional structural information encoded in images through physical imaging processes. The framework effectively preserves spatial geometric features including edges, hollow structures, and complex textures. It significantly enhances view-invariant reconstruction.Experiments on indoor scene reconstruction datasets confirm superior reconstruction stability. Our method maintains nearly identical quality when view count reduces by 40%. It achieves a coefficient of variation of 0.24%, performance retention rate of 99.7%, and maximum performance drop of 0.42%. These results demonstrate that exploiting intra-view spatial information provides a robust solution for view-limited scenarios in practical applications.

cs.CV↗

arXiv2025

ADP-VRSGP: Decentralized Learning with Adaptive Differential Privacy via Variance-Reduced Stochastic Gradient Push

Differential privacy is widely employed in decentralized learning to safeguard sensitive data by introducing noise into model updates. However, existing approaches that use fixed-variance noise often degrade model performance and reduce training efficiency. To address these limitations, we propose a novel approach called decentralized learning with adaptive differential privacy via variance-reduced stochastic gradient push (ADP-VRSGP). This method dynamically adjusts both the noise variance and the learning rate using a stepwise-decaying schedule, which accelerates training and enhances final model performance while providing node-level personalized privacy guarantees. To counteract the slowed convergence caused by large-variance noise in early iterations, we introduce a progressive gradient fusion strategy that leverages historical gradients. Furthermore, ADP-VRSGP incorporates decentralized push-sum and aggregation techniques, making it particularly suitable for time-varying communication topologies. Through rigorous theoretical analysis, we demonstrate that ADP-VRSGP achieves robust convergence with an appropriate learning rate, significantly improving training stability and speed. Experimental results validate that our method outperforms existing baselines across multiple scenarios, highlighting its efficacy in addressing the challenges of privacy-preserving decentralized learning.

cs.LG↗

arXiv2025

Post-training for Deepfake Speech Detection

We introduce a post-training approach that adapts self-supervised learning (SSL) models for deepfake speech detection by bridging the gap between general pre-training and domain-specific fine-tuning. We present AntiDeepfake models, a series of post-trained models developed using a large-scale multilingual speech dataset containing over 56,000 hours of genuine speech and 18,000 hours of speech with various artifacts in over one hundred languages. Experimental results show that the post-trained models already exhibit strong robustness and generalization to unseen deepfake speech. When they are further fine-tuned on the Deepfake-Eval-2024 dataset, these models consistently surpass existing state-of-the-art detectors that do not leverage post-training. Model checkpoints and source code are available online.

eess.AS↗

arXiv2025

Ontology-Enhanced Knowledge Graph Completion using Large Language Models

Large Language Models (LLMs) have been extensively adopted in Knowledge Graph Completion (KGC), showcasing significant research advancements. However, as black-box models driven by deep neural architectures, current LLM-based KGC methods rely on implicit knowledge representation with parallel propagation of erroneous knowledge, thereby hindering their ability to produce conclusive and decisive reasoning outcomes. We aim to integrate neural-perceptual structural information with ontological knowledge, leveraging the powerful capabilities of LLMs to achieve a deeper understanding of the intrinsic logic of the knowledge. We propose an ontology enhanced KGC method using LLMs -- OL-KGC. It first leverages neural perceptual mechanisms to effectively embed structural information into the textual space, and then uses an automated extraction algorithm to retrieve ontological knowledge from the knowledge graphs (KGs) that needs to be completed, which is further transformed into a textual format comprehensible to LLMs for providing logic guidance. We conducted extensive experiments on three widely-used benchmarks -- FB15K-237, UMLS and WN18RR. The experimental results demonstrate that OL-KGC significantly outperforms existing mainstream KGC methods across multiple evaluation metrics, achieving state-of-the-art performance.

cs.CL↗

arXiv2025

Fast Unbiased Sampling of Networks with Given Expected Degrees and Strengths

The configuration model is a cornerstone of statistical assessment of network structure. While the Chung-Lu model is among the most widely used configuration models, it systematically oversamples edges between large-degree nodes, leading to inaccurate statistical conclusions. Although the maximum entropy principle offers unbiased configuration models, its high computational cost has hindered widespread adoption, making the Chung-Lu model an inaccurate yet persistently practical choice. Here, we propose fast and efficient sampling algorithms for the max-entropy-based models by adapting the Miller-Hagberg algorithm. Evaluation on 103 empirical networks demonstrates 10-1000 times speedup, making theoretically rigorous configuration models practical and contributing to a more accurate understanding of network structure.

cs.SI↗

arXiv2025

Beyond Single Images: Retrieval Self-Augmented Unsupervised Camouflaged Object Detection

At the core of Camouflaged Object Detection (COD) lies segmenting objects from their highly similar surroundings. Previous efforts navigate this challenge primarily through image-level modeling or annotation-based optimization. Despite advancing considerably, this commonplace practice hardly taps valuable dataset-level contextual information or relies on laborious annotations. In this paper, we propose RISE, a RetrIeval SElf-augmented paradigm that exploits the entire training dataset to generate pseudo-labels for single images, which could be used to train COD models. RISE begins by constructing prototype libraries for environments and camouflaged objects using training images (without ground truth), followed by K-Nearest Neighbor (KNN) retrieval to generate pseudo-masks for each image based on these libraries. It is important to recognize that using only training images without annotations exerts a pronounced challenge in crafting high-quality prototype libraries. In this light, we introduce a Clustering-then-Retrieval (CR) strategy, where coarse masks are first generated through clustering, facilitating subsequent histogram-based image filtering and cross-category retrieval to produce high-confidence prototypes. In the KNN retrieval stage, to alleviate the effect of artifacts in feature maps, we propose Multi-View KNN Retrieval (MVKR), which integrates retrieval results from diverse views to produce more robust and precise pseudo-masks. Extensive experiments demonstrate that RISE outperforms state-of-the-art unsupervised and prompt-based methods. Code is available at https://github.com/xiaohainku/RISE.

cs.CV↗

arXiv2025

The Dual Nature of GHZ9: Coexisting Active Galactic Nuclei and Star Formation Activity in a Remote X-ray Source at z = 10.145

We present James Webb Space Telescope (JWST)/NIRSpec PRISM spectroscopic characterization of GHZ9 at z= 10.145 $\pm$ 0.010, currently the most distant source detected by the Chandra X-ray Observatory. The spectrum reveals several UV high-ionization lines, including CII, SiIV, NIV], CIV, HeII, OIII], NIII], and CIII]. The prominent rest-frame equivalent widths (EW(CIV)$\simeq$65A, EW(OIII])$\simeq$28A, EW(CIII])$\simeq$48A) show the presence of a hard active galactic nucleus (AGN) radiation field, while line ratio diagnostics are consistent with either AGN or star formation as the dominant ionizing source. GHZ9 is nitrogen-enriched (6--9.5 (N/O)$_{\odot}$), carbon-poor (0.2--0.65 (C/O)$_{\odot}$), metal-poor (Z = 0.01--0.1 Z$_{\odot}$), and compact ($<$ 106 pc), similarly to GN-z11, GHZ2, and recently discovered N-enhanced high redshift objects. We exploited the newly available JWST/NIRSpec and NIRCam data set to perform an independent analysis of the Chandra data confirming that GHZ9 is the most likely JWST source associated with X-ray emission at 0.5-7 keV. Assuming a spectral index $Γ$ = 2.3 (1.8), we estimate a black hole (BH) mass of 1.60 $\pm$ 0.31 (0.48 $\pm$ 0.09) $\times$ 10$^8$M$_{\odot}$, which is consistent either with Eddington-accretion onto heavy ($\geq$ 10$^6$ M$_{\odot}$) BH seeds formed at z=18, or super-Eddington accretion onto a light seed of $\sim$ 10$^2-10^4$ M$_{\odot}$ at z = 25. The corresponding BH-to-stellar mass ratio M$_{BH}$/M$_{star}$= 0.33$\pm$0.22 (0.10$\pm$0.07), with a stringent limit $>$0.02, implies an accelerated growth of the BH mass with respect to the stellar mass. GHZ9 is the ideal target to constrain the early phases of AGN-galaxy coevolution with future multi-frequency observations.

astro-ph.GA↗

arXiv2025

FakeMark: Deepfake Speech Attribution With Watermarked Artifacts

Deepfake speech attribution remains challenging for existing solutions. Classifier-based solutions often fail to generalize to domain-shifted samples, and watermarking-based solutions are easily compromised by distortions like codec compression or malicious removal attacks. To address these issues, we propose FakeMark, a novel watermarking framework that injects artifact-correlated watermarks associated with deepfake systems rather than pre-assigned bitstring messages. This design allows a detector to attribute the source system by leveraging both injected watermark and intrinsic deepfake artifacts, remaining effective even if one of these cues is elusive or removed. Experimental results show that FakeMark improves generalization to cross-dataset samples where classifier-based solutions struggle and maintains high accuracy under various distortions where conventional watermarking-based solutions fail.

eess.AS↗

arXiv2025

Quantum hypothesis testing via robust quantum control

Quantum hypothesis testing plays a pivotal role in quantum technologies, making decisions or drawing conclusions about quantum systems based on observed data. Recently, quantum control techniques have been successfully applied to quantum hypothesis testing, enabling the reduction of error probabilities in the task of distinguishing magnetic fields in presence of environmental noise. In real-world physical systems, such control is prone to various channels of inaccuracies. Therefore improving the robustness of quantum control in the context of quantum hypothesis testing is crucial. In this work, we utilize optimal control methods to compare scenarios with and without accounting for the effects of signal frequency inaccuracies. For parallel dephasing and spontaneous emission, the optimal control inherently demonstrates a certain level of robustness, while in the case of transverse dephasing with an imperfect signal, it may result in a higher error probability compared to the uncontrolled scheme. To overcome these limitations, we introduce a robust control approach optimized for a range of signal noise, demonstrating superior robustness beyond the predefined tolerance window. On average, both the optimal control and robust control show improvements over the uncontrolled schemes for various dephasing or decay rates, with the robust control yielding the lowest error probability.

quant-ph↗

arXiv2025

Jenga-Krotov algorithm: Efficient compilation of multi-qubit gates for exchange-only qubits

Exchange-only (EO) qubits, implemented in triple-quantum-dot systems, offer a compelling platform for scalable semiconductor-based quantum computing by enabling universal control through purely exchange interactions. While high-fidelity single- and two-qubit gates have been demonstrated, the synthesis of efficient multi-qubit operations-such as the Toffoli gate-remains a key bottleneck. Conventional gate decompositions into elementary operations lead to prohibitively long and error-prone pulse sequences, limiting practical deployment. In this work, we introduce a gradient-based optimization algorithm, Jenga-Krotov (JK), tailored to discover compact, high-fidelity EO gate sequences. Applying JK to the Toffoli gate, we reduce the number of required exchange unitaries from 216 (in direct decomposition) to 92, and compress the time steps required from 162 to 50, all while maintaining target fidelity. Under realistic noise, the accumulated gate error from our optimized sequence is an order of magnitude lower than that of conventional approaches. We have also applied the JK algorithm to other multi-qubit gates and algorithm. For the Fredkin gate, it reduces the number of time steps from 200 to 104 and the number of exchange unitaries from 276 to 172. For the quantum Fourier transform, it compresses the sequence from 180 to 80 time steps and from 237 to 202 exchange unitaries. These results demonstrate that the JK algorithm is a general and scalable strategy for multi-qubit gate synthesis in EO architectures, potentially facilitating realization of multi-qubit algorithms on semiconductor platforms.

quant-ph↗

arXiv2025

Hamiltonian Learning via Inverse Physics-Informed Neural Networks

Hamiltonian learning (HL), enabling precise estimation of system parameters and underlying dynamics, plays a critical role in characterizing quantum systems. However, conventional HL methods face challenges in noise robustness and resource efficiency, especially under limited measurements. In this work, we present \textit{Inverse Physics-Informed Neural Networks for Hamiltonian Learning (iPINN-HL)}, an approach that incorporates the Schrödinger equation as a soft constraint via a loss function penalty into the ML procedure. This formulation allows the model to integrate both observational data and known physical laws to infer Hamiltonian parameters with greater accuracy and resource efficiency. We benchmark iPINN-HL against a deep-neural-network-based quantum state tomography method (denoted as DNN-HL) and demonstrate its effectiveness across several different scenarios, including one-dimensional spin chains, cross-resonance gate calibration, crosstalk identification, and real-time compensation to parameter drift. Our results show that iPINN-HL can approach the Heisenberg limit and exhibits robustness to noises, while outperforming DNN-HL in accuracy and resource efficiency. Therefore, iPINN-HL is a powerful and flexible framework for quantum system characterization for practical tasks.

quant-ph↗

arXiv2025

Group Sparse Matrix Optimization for Efficient Quantum State Transformation

Finding ways to transform a quantum state to another is fundamental to quantum information processing. In this paper, we apply the sparse matrix approach to the quantum state transformation problem. In particular, we present a new approach for searching for unitary matrices for quantum state transformation by directly optimizing the objective problem using the Alternating Direction Method of Multipliers (ADMM). Moreover, we consider the use of group sparsity as an alternative sparsity choice in quantum state transformation problems. Our approach incorporates sparsity constraints into quantum state transformation by formulating it as a non-convex problem. It establishes a useful framework for efficiently handling complex quantum systems and achieving precise state transformations.

quant-ph↗

arXiv2025

Optimizing Quantum Transformation Matrices: A Block Decomposition Approach for Efficient Gate Reduction

This paper introduces an algorithm designed to approximate quantum transformation matrix with a restricted number of gates by using the block decomposition technique. Addressing challenges posed by numerous gates in handling large qubit transformations, the algorithm provides a solution by optimizing gate usage while maintaining computational accuracy. Inspired by the Block Decompose algorithm, our approach processes transformation matrices in a block-wise manner, enabling users to specify the desired gate count for flexibility in resource allocation. Simulations validate the effectiveness of the algorithm in approximating transformations with significantly fewer gates, enhancing quantum computing efficiency for complex calculations.

quant-ph↗

arXiv2025

Low-temperature series expansion of square lattice Ising model: A study based on Fisher zeros

Low-temperature expansion of Ising model has long been a topic of significant interest in condensed matter and statistical physics. In this paper we present new results of the coefficients in the low-temperature series of the Ising partition function on the square lattice, in the cases of a zero field and of an imaginary field $i(π/2)k_BT$. The coefficients in the low-temperature series of the free energy in the thermodynamic limit are represented using the explicit expression of the density function of the Fisher zeros. The asymptotic behaviour of the sequence of the coefficients when the order goes to infinity is determined exactly, for both the series of the free energy and of the partition function. Our analytic and numerical results demonstrate that, the convergence radius of the sequence is dependent on the accumulation points of the Fisher zeros which have the smallest modulus. In the zero field case this accumulation point is the physical critical point, while in the imaginary field case it corresponds to a non-physical singularity. We further discuss the relation between the series coefficients and the energy state degeneracies, using the combinatorial expression of the coefficients and the subgraph expansion.

cond-mat.stat-mech↗