arXiv subjects

Ran Li

Publications and source records attributed to Ran Li.

At least 109 records · Page 6Linked to original sources

arXiv2025

REAL: Realism Evaluation of Text-to-Image Generation Models for Effective Data Augmentation

Recent advancements in text-to-image (T2I) generation models have transformed the field. However, challenges persist in generating images that reflect demanding textual descriptions, especially for fine-grained details and unusual relationships. Existing evaluation metrics focus on text-image alignment but overlook the realism of the generated image, which can be crucial for downstream applications like data augmentation in machine learning. To address this gap, we propose REAL, an automatic evaluation framework that assesses realism of T2I outputs along three dimensions: fine-grained visual attributes, unusual visual relationships, and visual styles. REAL achieves a Spearman's rho score of up to 0.62 in alignment with human judgement and demonstrates utility in ranking and filtering augmented data for tasks like image captioning, classification, and visual relationship detection. Empirical results show that high-scoring images evaluated by our metrics improve F1 scores of image classification by up to 11.3%, while low-scoring ones degrade that by up to 4.95%. We benchmark four major T2I models across the realism dimensions, providing insights for future improvements in T2I output realism.

cs.MM↗

arXiv2024

Note on the local calculation of decoherence of quantum superpositions in de Sitter spacetime

We study the decoherence effect of quantum superposition in de Sitter (dS) spacetime due to the presence of the cosmological horizon. Using the algebraic approach of quantum field theory on curved spacetime, we derive the precise expression for the expected number of entangling particles in the scalar field case. This expression establishes the relation between the decoherence and the local two-point correlation function. Specifically, we analyze the quantum superposition Gendankenexperiment performed by a local observer at the center of dS spacetime. We compute the entangling particle numbers in scalar field, electromagnetic field, and gravitational field scenarios. It is demonstrated that the quantum spatial superposition state can be decohered by emitting entangling particles into the cosmological horizon. Our setup is equivalent to an accelerating observer in 5-dimensional Minkowski spacetime. The results for the scalar and electromagnetic cases are consistent with those obtained in Ref.[1], which investigated the decoherence effect from the perspective of an accelerating observer in Minkovski spacetime. However, our result fixes the numerical prefactor of the gravitational decoherence.

hep-th↗

arXiv2024

Uniform measure attractors of McKean-Vlasov stochastic reaction-diffusion equations on unbounded thin domain

This article addresses the issue of uniform measure attractors for non-autonomous McKean-Vlasov stochastic reaction-diffusion equations defined on unbounded thin domains. Initially, the concept of uniform measure attractors is recalled, and thereafter, the existence and uniqueness of such attractors are demonstrated. Uniform tail estimates are employed to establish the asymptotic compactness of the processes, thereby overcoming the non-compactness issue inherent in the usual Sobolev embedding on unbounded thin domains. Finally, we demonstrate that the upper semi-continuity of uniform measure attractors defined on $(n + 1)$-dimensional unbounded thin domains collapsing into the space $\mathbb{R}^n$.

math.PR↗

arXiv2024

Decoherence of quantum superpositions by Reissner-Nordström black holes

Recently, it was shown by Danielson-Satishchandran-Wald (DSW) that for the massive or charged body in a quantum spatial separated superposition state, the presence of a black hole can decohere the superposition inevitably towards capturing the radiation of soft photons or gravitons. In this work, we study the DSW decoherence effect for the static charged body in the Reissner-Nordström black holes. By calculating the decohering rate for this case, it is shown that the superposition is decohered by the low frequency photons that propagate through the black hole horizon. For the extremal Reissner-Nordström black hole, the decoherence of quantum superposition is completely suppressed due to the black hole Meissner effect.It is also found that the decoherence effect caused by the Reissner-Nordström black hole is equivalent to that of an ordinary matter system with the same size and charge.

hep-th↗

arXiv2024

Long-time Behaviour of the Non-autonomous Stochastic FitzHugh-Nagumo Systems on Thin Domains

We study the long-time behavior of non-autonomous stochastic FitzHugh-Nagumo systems on thin domains. As the $(n+ 1)$-dimensional thin domains collapses onto an n-dimensional domain, an n-dimensional limiting FitzHugh-Nagumo system is derived. This n-dimensional limiting system encodes the defining geometry of the $(n+1)$-dimensional system. To justify this limiting process, we show that the pullback measure attractors of the FitzHugh-Nagumo systems on thin domains are upper semi-continuous as the height of thin direction tends to zero.

math.DS↗

arXiv2024

FedQP: Towards Accurate Federated Learning using Quadratic Programming Guided Mutation

Due to the advantages of privacy-preserving, Federated Learning (FL) is widely used in distributed machine learning systems. However, existing FL methods suffer from low-inference performance caused by data heterogeneity. Specifically, due to heterogeneous data, the optimization directions of different local models vary greatly, making it difficult for the traditional FL method to get a generalized global model that performs well on all clients. As one of the state-of-the-art FL methods, the mutation-based FL method attempts to adopt a stochastic mutation strategy to guide the model training towards a well-generalized area (i.e., flat area in the loss landscape). Specifically, mutation allows the model to shift within the solution space, providing an opportunity to escape areas with poor generalization (i.e., sharp area). However, the stochastic mutation strategy easily results in diverse optimal directions of mutated models, which limits the performance of the existing mutation-based FL method. To achieve higher performance, this paper proposes a novel mutation-based FL approach named FedQP, utilizing a quadratic programming strategy to regulate the mutation directions wisely. By biasing the model mutation towards the direction of gradient update rather than traditional random mutation, FedQP can effectively guide the model to optimize towards a well-generalized area (i.e., flat area). Experiments on multiple well-known datasets show that our quadratic programming-guided mutation strategy effectively improves the inference accuracy of the global model in various heterogeneous data scenarios.

cs.LG↗

arXiv2024

UPdec-Webb: A Dataset for Coaddition of JWST NIRCam Images

We present the application of the image coaddition algorithm, Up-sampling and PSF Deconvolution Coaddition (UPDC), for stacking multiple exposure images captured by the James Webb Space Telescope (JWST) Near-Infrared Camera (NIRCam). By addressing the point spread function (PSF) effect, UPDC provides visually enhanced and sharper images. Furthermore, the anti-aliasing and super-resolution capabilities of UPDC make it easier to deblend sources overlapped on images, yielding a higher accuracy of aperture photometry. We apply this algorithm to the SMACS J0723 imaging data. Comparative analysis with the Drizzle algorithm demonstrates significant improvements in detecting faint sources, achieving accurate photometry, and effectively deblending (super-resolution) closely packed sources. {As a result, we have newly detected a pair of close binary stars that were previously unresolvable in the original exposures or the Drizzled image.} These improvements significantly benefit various scientific projects conducted by JWST. The resulting dataset, named "UPdec-Webb", can be accessible through the official website of the Chinese Virtual Observatory (ChinaVO).

astro-ph.IM↗

arXiv2024

RAFT: Realistic Attacks to Fool Text Detectors

Large language models (LLMs) have exhibited remarkable fluency across various tasks. However, their unethical applications, such as disseminating disinformation, have become a growing concern. Although recent works have proposed a number of LLM detection methods, their robustness and reliability remain unclear. In this paper, we present RAFT: a grammar error-free black-box attack against existing LLM detectors. In contrast to previous attacks for language models, our method exploits the transferability of LLM embeddings at the word-level while preserving the original text quality. We leverage an auxiliary embedding to greedily select candidate words to perturb against the target detector. Experiments reveal that our attack effectively compromises all detectors in the study across various domains by up to 99%, and are transferable across source models. Manual human evaluation studies show our attacks are realistic and indistinguishable from original human-written text. We also show that examples generated by RAFT can be used to train adversarially robust detectors. Our work shows that current LLM detectors are not adversarially robust, underscoring the urgent need for more resilient detection mechanisms.

cs.CL↗

arXiv2024

Thermodynamics and kinetics of state switching for the asymptotically flat black hole in a cavity

We propose that the thermodynamics and the kinetics of state switching for the asymptotically flat black hole enclosed by a cavity can be studied in terms of the free energy landscape formalism. The generalized free energy for the black hole enclosed by a cavity in the canonical ensemble is derived by using the York's approach, where the temperature on the cavity and the charges inside the cavity are kept as the fixed parameters. By quantifying the corresponding free energy landscape, we obtain the phase diagrams for the black hole in cavity, which reveal a Hawking-Page type transition for the uncharged black hole and a Van der Waals type transition for the charged black hole. We further assume that the dynamics of black hole state switching is mutually determined by the gradient force and the stochastic force arising from the free energy landscape and the thermal noises respectively. We then derive a recurrence relation for the $n$-momentum of the first passage time distribution function, which enables the calculation of the kinetic times characterized by the mean first passage time and its relative fluctuation. Our analysis illustrates that the kinetics of black hole state switching is determined by the ensemble temperature and the barrier height on the free energy landscape.

gr-qc↗

arXiv2024

Preparation for CSST: Star-galaxy Classification using a Rotationally Invariant Supervised Machine Learning Method

Most existing star-galaxy classifiers depend on the reduced information from catalogs, necessitating careful data processing and feature extraction. In this study, we employ a supervised machine learning method (GoogLeNet) to automatically classify stars and galaxies in the COSMOS field. Unlike traditional machine learning methods, we introduce several preprocessing techniques, including noise reduction and the unwrapping of denoised images in polar coordinates, applied to our carefully selected samples of stars and galaxies. By dividing the selected samples into training and validation sets in an 8:2 ratio, we evaluate the performance of the GoogLeNet model in distinguishing between stars and galaxies. The results indicate that the GoogLeNet model is highly effective, achieving accuracies of 99.6% and 99.9% for stars and galaxies, respectively. Furthermore, by comparing the results with and without preprocessing, we find that preprocessing can significantly improve classification accuracy (by approximately 2.0% to 6.0%) when the images are rotated. In preparation for the future launch of the China Space Station Telescope (CSST), we also evaluate the performance of the GoogLeNet model on the CSST simulation data. These results demonstrate a high level of accuracy (approximately 99.8%), indicating that this model can be effectively utilized for future observations with the CSST.

astro-ph.GA↗

arXiv2024

The Velocity Aberration Effect of the CSST Main Survey Camera

In this study, we conducted simulations to find the geometric aberrations expected for images taken by the Main Survey Camera (MSC) of the Chinese Space Station Telescope (CSST) due to its motion. As anticipated by previous work, our findings indicate that the geometric distortion of light impacts the focal plane's apparent scale, with a more pronounced influence as the size of the focal plane increases. Our models suggest that the effect consistently influences the pixel scale in both the vertical and parallel directions. The apparent scale variation follows a sinusoidal distribution throughout one orbit period. Simulations reveal that the effect is particularly pronounced in the center of the Galaxy and gradually diminishes along the direction of ecliptic latitude. At low ecliptic latitudes, the total aberration leads to about 0.94 pixels offset (a 20-minute exposure) and 0.26 pixels offset (a 300-second exposure) at the edge of the field of view, respectively. Appropriate processings for the geometric effect during the CSST pre- and post-observation phases are presented.

astro-ph.IM↗

arXiv2024

Multi-band analysis of strong gravitationally lensed post-blue nugget candidates from the Kilo-Degree Survey

During the early stages of galaxy evolution, a significant fraction of galaxies undergo a transitional phase between the ``blue nugget" systems, which arise from the compaction of large, active star-forming disks, and the ``red nuggets", which are red and passive compact galaxies. These objects are typically only observable with space telescopes, and detailed studies of their size, mass, and stellar population parameters have been conducted on relatively small samples. Strong gravitational lensing can offer a new opportunity to study them in detail, even with ground-based observations. In this study, we present the first 6 \textit{bona fide} sample of strongly lensed post-blue nugget (pBN) galaxies, which were discovered in the Kilo Degree Survey. By using the lensing-magnified luminosity from optical and near-infrared bands, we have derived robust structural and stellar population properties of the multiple images of the background sources. \rui{The pBN galaxies have very small sizes of $\rm R_{eff}<1.3$ kpc, high mass density inside 1 kpc of $ \rm \log (Σ_1/M_{\odot} \mathrm{kpc}^{-2})>9.3$, and low specific star formation rates of $\log (\mathrm{sSFR/Gyr})\lesssim0$, The size-mass and $Σ_1$-mass relations of this sample are consistent with those of the red nuggets, while their sSFR is close to the lower end of compact star-forming blue nugget systems at the same redshift, suggesting a clear evolutionary link between them.

astro-ph.GA↗

arXiv2024

CSST Strong Lensing Preparation: Forecasting the galaxy-galaxy strong lensing population for the China Space Station Telescope

Galaxy-galaxy strong gravitational lensing (GGSL) is a powerful probe for the formation and evolution of galaxies and cosmology, while the sample size of GGSLs leads to considerable uncertainties and potential bias. The China Space Station Telescope (CSST, to be launched in late 2026) will conduct observations across 17,500 square degrees of the sky, capturing images in the $ugriz$ bands with a spatial resolution comparable to that of the Hubble Space Telescope. We ran a set of Monte Carlo simulations to predict that the CSST's wide-field survey will observe $\sim$160,000 galaxy-galaxy strong lenses over its lifespan, increasing the number of existing galaxy-galaxy strong lens samples by three orders of magnitude. This is comparable to the capabilities of the $\it Euclid$ telescope but with the added benefit of additional color information. Specifically, the CSST can detect strong lenses with Einstein radii about $0.64\pm0.42^{"}$, corresponding to the velocity dispersions of $217.19 \pm 50.55 \, \text{km/s}$. These lenses exhibit a median magnification of $\sim$5. The apparent magnitude of the unlensed sources in the g-band is $25.87 \pm 1.19$. The signal-to-noise ratio of the lensed images covers a range of $\sim 20$ to $\sim 1000$, allowing us to determine the Einstein radius with an accuracy ranging from $\sim 1 \%$ to $\sim 0.1 \%$, ignoring various modeling systematics. Our estimates indicate that CSST can observe rare systems like double source-plane and spiral galaxy lenses. The above selection functions of the CSST strong lensing observation help optimize the strategy of finding and modeling GGSLs.

astro-ph.GA↗

arXiv2024

Internet of Agents: Weaving a Web of Heterogeneous Agents for Collaborative Intelligence

The rapid advancement of large language models (LLMs) has paved the way for the development of highly capable autonomous agents. However, existing multi-agent frameworks often struggle with integrating diverse capable third-party agents due to reliance on agents defined within their own ecosystems. They also face challenges in simulating distributed environments, as most frameworks are limited to single-device setups. Furthermore, these frameworks often rely on hard-coded communication pipelines, limiting their adaptability to dynamic task requirements. Inspired by the concept of the Internet, we propose the Internet of Agents (IoA), a novel framework that addresses these limitations by providing a flexible and scalable platform for LLM-based multi-agent collaboration. IoA introduces an agent integration protocol, an instant-messaging-like architecture design, and dynamic mechanisms for agent teaming and conversation flow control. Through extensive experiments on general assistant tasks, embodied AI tasks, and retrieval-augmented generation benchmarks, we demonstrate that IoA consistently outperforms state-of-the-art baselines, showcasing its ability to facilitate effective collaboration among heterogeneous agents. IoA represents a step towards linking diverse agents in an Internet-like environment, where agents can seamlessly collaborate to achieve greater intelligence and capabilities. Our codebase has been released at \url{https://github.com/OpenBMB/IoA}.

cs.CL↗

arXiv2024

Entanglement island and Page curve of Hawking radiation in rotating Kerr black holes

We initiate the study of the information paradox of rotating Kerr black holes by employing the recently proposed island rule. It is known that the scalar field theory near the Kerr black hole horizon can be reduced to the 2-dimensional effective theory. Working within the framework of the 2-dimensional effective theory and assuming the small angular momentum limit, we demonstrate that the entanglement entropy of Hawking radiation from the non-extremal Kerr black hole follows the Page curve and saturates the Bekenstein-Hawking entropy at late times. In addition, we also discuss the effect of the black hole rotation on the Page time and scrambling time. For the extreme Kerr black hole, the entanglement entropy at late times also approximates the Bekenstein-Hawking entropy of the extreme Kerr black hole. These results imply that entanglement islands can provide a semi-classical resolution of the information paradox for rotating Kerr black holes.

hep-th↗

arXiv2024

On the fidelity of information retrieval in the black hole final state model with scrambling interactions

We study the fidelity of information retrieval in the black hole final state model by taking into account the interactions between the collapsing matter and the infalling Hawking radiation inside the event horizon. By utilizing a scrambling unitary operator to model these interactions, our direct calculations suggest that the information is almost lost during the process of black hole evaporation. We then proceed to employ the Yoshida-Kitaev decoding strategy for information retrieval. Although we observe a finite improvement in decoding fidelity, it does not reach unity. This indicates that the issue of unitarity in black hole evaporation may not be fully resolved within the framework of the final state model.

hep-th↗

arXiv2024

The measurement of the splashback radius of dark matter halo

In the hierarchical evolution framework of cosmology, larger halos grow through matter accretion and halo mergers. To clarify the halo evolution, we need to define the halo mass and radius physically. However, the pseudo-evolution problem makes the process difficult. Thus, we aim to measure the splashback radius, a physically defined halo radius for a large number of halos with various mass and redshift, and to determine the most important parameters to affect it. We use the typical definition of splashback radius (Rsp) as the radius with the steepest radial density profile. In this work, we measure Rsp of dark matter halos within the mass of 1e13-3e15Msun and redshifts spanning 0.08-0.65. This is the measurement of the Rsp in the largest range of halo mass and redshift. Using the shear catalog of the DECaLS DR8, we investigate Rsp of halos associated with galaxies and galaxy clusters identified in the various catalogs. Our finding reveals a trend wherein massive halos demonstrate a larger Rsp, and the normalized splashback radius (Rsp/R200m) shows a U-shaped mass evolution. The upturn in these relations mainly comes from the contribution of massive halos with low redshifts. We further find Rsp increases with the peak height, while Rsp/R200m has a negative relation with the peak height. We also find the Rsp >~R200m for most halos, indicating their low accretion rates. Our result is consistent with previous literature across a wide range of mass, redshift, and peak height, as well as the simulation work from More et al. (2015).

astro-ph.CO↗

arXiv2024

The XMM-SERVS X-ray eXtended Galaxy Cluster (XVXGC) catalog

To explain the well-known tension between cosmological parameter constraints obtained from the primary CMB and those drawn from galaxy cluster samples, we propose a possible explanation for the incompleteness of detected clusters are higher than estimated. We aim to search for galaxy groups and clusters with particularly extended surface brightness distributions by creating a new X-ray-selected catalog of extended galaxy clusters from the XMM-SERVS data, based on a dedicated source detection and characterization algorithm that is optimized for extended sources. Our state-of-the-art algorithm is composed of wavelet filtering, source detection, and characterization. We make a visual inspection of the optical image, and spatial distribution of galaxies within the same redshift layer to confirm the existence of clusters and estimate the cluster redshift with the spectroscopic and photometric redshifts of galaxies. The growth curve analysis is used to characterize the detections. We report a catalog of extended X-ray galaxy clusters detected from the XMM-SERVS data, named the XMM- SERVS X-ray eXtended Galaxy Cluster (XVXGC) catalog. It includes 141 cluster candidates. Specifically, there are 52 clusters previously identified as clusters with the intra-cluster medium (ICM) emission (class 3), 37 ones previously known as optical or infrared clusters but detected as X-ray clusters for the first time (class 2), and 52 identified as clusters for the first time (class 1). Compared with the class3 sample, the 'class1+2' sample is systematically fainter, and exhibits a flatter surface brightness profile. The median flux in [0.1-2.4]keV band for 'class1+2' and class3 sample is 2.336e-14 and 3.163e-14erg/s/cm2, respectively. The median slope of surface brightness profile are 0.502 and 0.577 for the 'class1+2' and class 3 samples, respectively.

astro-ph.CO↗