arXiv subjects

Wei Du

Publications and source records attributed to Wei Du.

At least 73 records · Page 4Linked to original sources

arXiv2024

Study of Group III-V Waveguides on Sapphire Platform for Photonic Integrated Circuits

Photonic integrated circuits (PICs) have been acknowledged as the promising platforms for the applications in data communication, Lidar in autonomous driving vehicles, innovative sensor technology, etc. Since the demonstration of optical components individually, integration of both electronics and photonics for functional devices on a common platform has been a key technology driver enhancing the stability and scalability of integrated photonic technologies. Recently, we proposed to use sapphire as a high-performance PIC platform, which enables a fully integrated solution to include a complete set of components with light source, modulator, light detection, passive devices, silicon on sapphire control circuit all-in-one sapphire platform to achieve high-performance low-cost mixed-signal optical links. In parallel to developing ac-tive components such as group III-V lasers on sapphire, in this work, the performance of group III-V straight waveguides on sapphire was systemically studied. The refractive indices contrast between GaAs, InP, GaSb, and sapphire are sufficiently high to achieve low loss over a broad optical wavelength. The calculated loss at wavelengths of 1330 nm, 1550 nm, and 2000 nm for the GaAs, InP, and GaSb rib waveguides are 0.32 dB/cm, 0.67 dB/cm, and 0.70 dB/cm, re-spectively. Since the fundamental element to construct all passive building blocks is the straight waveguide, results from this work would allow us to assess other basic passive building blocks.

physics.optics↗

arXiv2024

Quantum Locking of Intrinsic Spin Squeezed State in Earth-field-range Magnetometry

In the Earth-field range, the nonlinear Zeeman (NLZ) effect has been a bottleneck limiting the sensitivity and accuracy of atomic magnetometry from physical mechanism. To break this bottleneck, various techniques are introduced to suppress the NLZ effect. Here we revisit the spin dynamics in the Earth-field-range magnetometry and identify the existence of the intrinsic spin squeezed state (SSS) generated from the geomagnetically induced NLZ effect with the oscillating squeezing degree and squeezing axis. Such oscillating features of the SSS prevent its direct observation and as well, accessibility to magnetic sensing. To exploit quantum advantage of the intrinsic SSS in the Earth-field-range magnetometry, it's essential to lock the oscillating SSS to a persistent one. Hence, we develop a quantum locking technique to achieve a persistent SSS, benefiting from which the sensitivity of the Earth-field-range magnetometer is quantum-enhanced. This work presents an innovative way turning the drawback of NLZ effect into the quantum advantage and opens a new access to quantum-enhanced magnetometry in the Earth-field range.

physics.atom-ph↗

arXiv2024

Photometric Redshift Estimation for CSST Survey with LSTM Neural Networks

Accurate estimation of photometric redshifts (photo-$z$s) is crucial for cosmological surveys. Various methods have been developed for this purpose, such as template fitting methods and machine learning techniques, each with its own applications, advantages, and limitations. In this study, we propose a new approach that utilizes a deep learning model based on Recurrent Neural Networks (RNN) with Long Short-Term Memory (LSTM) to predict photo-$z$. Unlike many existing machine learning models, our method requires only flux measurements from different observed filters as input. The model can automatically learn the complex relationships between the flux data across different wavelengths, eliminating the need for manually extracted or derived input features, thereby providing precise photo-$z$ estimates. The effectiveness of our proposed model is evaluated using simulated data from the Chinese Space Station Telescope (CSST) sourced from the Hubble Space Telescope Advanced Camera for Surveys (HST-ACS) and the COSMOS catalog, considering anticipated instrument effects of the future CSST. Results from experiments demonstrate that our LSTM model, compared to commonly used template fitting and machine learning approaches, requires minimal input parameters and achieves high precision in photo-$z$ estimation. For instance, when trained on the same dataset and provided only with photometric fluxes as input features, the proposed LSTM model yields one-third of the outliers $f_{out}$ observed with a Multi-Layer Perceptron Neural Network (MLP) model, while the normalized median absolute deviation $\rm σ_{NMAD}$ is only two-thirds that of the MLP model. This study presents a novel approach to accurately estimate photo-$z$s of galaxies using photometric data from large-scale survey projects.

astro-ph.GA↗

arXiv2024

Modeling the wavelength dependence of pixel response non-uniformity of a CCD sensor

Precision measurements in astronomy require stringent control of systematics such as those arising from imperfect correction of sensor effects. In this work, we develop a parametric method to model the wavelength dependence of pixel response non-uniformity (PRNU) for a laser-annealed backside-illuminated charge-coupled device. The model accurately reproduces the PRNU patterns of flat-field images taken at nine wavelengths from 290nm to 950nm, leaving the root mean square (RMS) residuals no more than 0.2% in most cases. By removing the large-scale non-uniformity in the flat fields, the RMS residuals are further reduced. This model fitting approach gives more accurate predictions of the PRNU than cubic-spline interpolation does with fewer free parameters. It can be applied to make PRNU corrections for individual objects according to their spectral energy distribution to reduce the photometry errors caused by the wavelength-dependent PRNU, if sub-percent level precision is required.

astro-ph.IM↗

arXiv2024

OpenMathInstruct-2: Accelerating AI for Math with Massive Open-Source Instruction Data

Mathematical reasoning continues to be a critical challenge in large language model (LLM) development with significant interest. However, most of the cutting-edge progress in mathematical reasoning with LLMs has become \emph{closed-source} due to lack of access to training data. This lack of data access limits researchers from understanding the impact of different choices for synthesizing and utilizing the data. With the goal of creating a high-quality finetuning (SFT) dataset for math reasoning, we conduct careful ablation experiments on data synthesis using the recently released \texttt{Llama3.1} family of models. Our experiments show that: (a) solution format matters, with excessively verbose solutions proving detrimental to SFT performance, (b) data generated by a strong teacher outperforms equally-sized data generated by a weak student model, (c) SFT is robust to low-quality solutions, allowing for imprecise data filtering, and (d) question diversity is crucial for achieving data scaling gains. Based on these insights, we create the OpenMathInstruct-2 dataset, which consists of 14M question-solution pairs ($\approx$ 600K unique questions), making it nearly eight times larger than the previous largest open-source math reasoning dataset. Finetuning the \texttt{Llama-3.1-8B-Base} using OpenMathInstruct-2 outperforms \texttt{Llama3.1-8B-Instruct} on MATH by an absolute 15.9\% (51.9\% $\rightarrow$ 67.8\%). Finally, to accelerate the open-source efforts, we release the code, the finetuned models, and the OpenMathInstruct-2 dataset under a commercially permissive license.

cs.CL↗

arXiv2024

Towards reliable respiratory disease diagnosis based on cough sounds and vision transformers

Recent advancements in deep learning techniques have sparked performance boosts in various real-world applications including disease diagnosis based on multi-modal medical data. Cough sound data-based respiratory disease (e.g., COVID-19 and Chronic Obstructive Pulmonary Disease) diagnosis has also attracted much attention. However, existing works usually utilise traditional machine learning or deep models of moderate scales. On the other hand, the developed approaches are trained and evaluated on small-scale data due to the difficulty of curating and annotating clinical data on scale. To address these issues in prior works, we create a unified framework to evaluate various deep models from lightweight Convolutional Neural Networks (e.g., ResNet18) to modern vision transformers and compare their performance in respiratory disease classification. Based on the observations from such an extensive empirical study, we propose a novel approach to cough-based disease classification based on both self-supervised and supervised learning on a large-scale cough data set. Experimental results demonstrate our proposed approach outperforms prior arts consistently on two benchmark datasets for COVID-19 diagnosis and a proprietary dataset for COPD/non-COPD classification with an AUROC of 92.5%.

cs.CV↗

arXiv2024

Transferring Backdoors between Large Language Models by Knowledge Distillation

Backdoor Attacks have been a serious vulnerability against Large Language Models (LLMs). However, previous methods only reveal such risk in specific models, or present tasks transferability after attacking the pre-trained phase. So, how risky is the model transferability of a backdoor attack? In this paper, we focus on whether existing mini-LLMs may be unconsciously instructed in backdoor knowledge by poisoned teacher LLMs through knowledge distillation (KD). Specifically, we propose ATBA, an adaptive transferable backdoor attack, which can effectively distill the backdoor of teacher LLMs into small models when only executing clean-tuning. We first propose the Target Trigger Generation (TTG) module that filters out a set of indicative trigger candidates from the token list based on cosine similarity distribution. Then, we exploit a shadow model to imitate the distilling process and introduce an Adaptive Trigger Optimization (ATO) module to realize a gradient-based greedy feedback to search optimal triggers. Extensive experiments show that ATBA generates not only positive guidance for student models but also implicitly transfers backdoor knowledge. Our attack is robust and stealthy, with over 80% backdoor transferability, and hopes the attention of security.

cs.CR↗

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

TrojanRAG: Retrieval-Augmented Generation Can Be Backdoor Driver in Large Language Models

Large language models (LLMs) have raised concerns about potential security threats despite performing significantly in Natural Language Processing (NLP). Backdoor attacks initially verified that LLM is doing substantial harm at all stages, but the cost and robustness have been criticized. Attacking LLMs is inherently risky in security review, while prohibitively expensive. Besides, the continuous iteration of LLMs will degrade the robustness of backdoors. In this paper, we propose TrojanRAG, which employs a joint backdoor attack in the Retrieval-Augmented Generation, thereby manipulating LLMs in universal attack scenarios. Specifically, the adversary constructs elaborate target contexts and trigger sets. Multiple pairs of backdoor shortcuts are orthogonally optimized by contrastive learning, thus constraining the triggering conditions to a parameter subspace to improve the matching. To improve the recall of the RAG for the target contexts, we introduce a knowledge graph to construct structured data to achieve hard matching at a fine-grained level. Moreover, we normalize the backdoor scenarios in LLMs to analyze the real harm caused by backdoors from both attackers' and users' perspectives and further verify whether the context is a favorable tool for jailbreaking models. Extensive experimental results on truthfulness, language understanding, and harmfulness show that TrojanRAG exhibits versatility threats while maintaining retrieval capabilities on normal queries.

cs.CR↗

arXiv2024

Acceleration of the Universe without the Hubble tension with Kaniadakis holographic dark energy using the Hubble horizon as the IR cut-off

We introduce a holographic dark energy model that incorporates the first-order approximate Kaniadaski entropy, utilizing the Hubble horizon, $1/H$, as the infrared cutoff. We investigate the cosmological evolution within this framework. The model introduces an extra parameter relative to the $Λ$CDM model. It posits a Universe that is initially dominated by dark matter, which then evolves to a phase where dark energy becomes the predominant component, with this transition occurring at a redshift of approximately $z \sim 0.419$. The energy density of dark energy is ultimately expected to become constant, thereby circumventing the potential issue of a "big rip". Employing the most recent Type Ia supernova and Hubble parameter data, we constrain the model's parameters and find a Hubble constant of $H_0=72.8$ km/s/Mpc, thereby resolving the Hubble tension issue. The estimated age of the Universe, based on the best-fit parameter values, is $14.2$ Gyr. Furthermore, we predict the number of strong gravitational lenses and conduct statefinder and $Om$ diagnostic analyses to validate and characterize the model.

astro-ph.CO↗

arXiv2024

Imputation of Missing Photometric Data and Photometric Redshift Estimation for CSST

Accurate photometric redshift (photo-$z$) estimation requires support from multi-band observational data. However, in the actual process of astronomical observations and data processing, some sources may have missing observational data in certain bands for various reasons. This could greatly affect the accuracy and reliability of photo-$z$ estimation for these sources, and even render some estimation methods unusable. The same situation may exist for the upcoming Chinese Space Station Telescope (CSST). In this study, we employ a deep learning method called Generative Adversarial Imputation Networks (GAIN) to impute the missing photometric data in CSST, aiming to reduce the impact of data missing on photo-$z$ estimation and improve estimation accuracy. Our results demonstrate that using the GAIN technique can effectively fill in the missing photometric data in CSST. Particularly, when the data missing rate is below 30\%, the imputation of photometric data exhibits high accuracy, with higher accuracy in the $g$, $r$, $i$, $z$, and $y$ bands compared to the $NUV$ and $u$ bands. After filling in the missing values, the quality of photo-$z$ estimation obtained by the widely used Easy and Accurate Zphot from Yale (EAZY) software is notably enhanced. Evaluation metrics for assessing the quality of photo-$z$ estimation, including the catastrophic outlier fraction ($f_{out}$), the normalized median absolute deviation ($\rm {σ_{NMAD}}$), and the bias of photometric redshift ($bias$), all show some degree of improvement. Our research will help maximize the utilization of observational data and provide a new method for handling sample missing values for applications that require complete photometry data to produce results.

astro-ph.IM↗

arXiv2024

Investigating Multi-Hop Factual Shortcuts in Knowledge Editing of Large Language Models

Recent work has showcased the powerful capability of large language models (LLMs) in recalling knowledge and reasoning. However, the reliability of LLMs in combining these two capabilities into reasoning through multi-hop facts has not been widely explored. This paper systematically investigates the possibilities for LLMs to utilize shortcuts based on direct connections between the initial and terminal entities of multi-hop knowledge. We first explore the existence of factual shortcuts through Knowledge Neurons, revealing that: (i) the strength of factual shortcuts is highly correlated with the frequency of co-occurrence of initial and terminal entities in the pre-training corpora; (ii) few-shot prompting leverage more shortcuts in answering multi-hop questions compared to chain-of-thought prompting. Then, we analyze the risks posed by factual shortcuts from the perspective of multi-hop knowledge editing. Analysis shows that approximately 20% of the failures are attributed to shortcuts, and the initial and terminal entities in these failure instances usually have higher co-occurrences in the pre-training corpus. Finally, we propose erasing shortcut neurons to mitigate the associated risks and find that this approach significantly reduces failures in multiple-hop knowledge editing caused by shortcuts.

cs.CL↗

arXiv2024

Monolithic Germanium Tin on Si Avalanche Photodiodes

We demonstrate monolithically grown germanium-tin (GeSn) on silicon avalanche photodiodes (APDs) for infrared light detection. A relatively thinner Ge buffer design was adopted to allow effective photo carriers to transport from the GeSn absorber to the Si multiplication layer such that clear punch-through behavior and a saturated primary responsivity of 0.3 A/W at 1550 nm were observed before avalanche breakdown in GeSn/Si APDs for the first time. The spectral response covers 1500 to 1700 nm. The measured punch-through and breakdown voltages are 15 and 17 V, respectively. Undisputed multiplication gain was obtained with the maximum value of 4.5 at 77 K, and 1.4 at 250 K, directly in reference to the saturated primary responsivity from the same device rather than a different GeSn p-i-n photodiode in previous reports. A peak responsivity was measured as 1.12 A/W at 1550 nm and 77 K.

cond-mat.mtrl-sci↗

arXiv2024

Electrically Injected mid-infrared GeSn laser on Si operating at 140 K

Owing to its true direct bandgap and tunable bandgap energies,GeSn alloys are increasingly attractive as gain media for mid-IR lasers that can be monolithically integrated on Si. Demonstrations of optically pumped GeSn laser at room under pulsed condition and at cryogenic temperature under continuous-wave excitation show great promise of GeSn lasers to be efficient electrically injected light sources on Si. Here we report electrically injected GeSn lasers using Fabry-Perot cavity with 20, 40, and 80 micron ridge widths. A maximum operating temperature of 140 K with lasing threshold of 0.756 kA/cm2 at 77 K and emitting wavelength of 2722 nm at 140 K was obtained. The lower threshold current density compared to previous works was achieved by reducing optical loss and improving the optical confinement. The peak power was measured as 2.2 mW/facet at 77 K.

physics.optics↗

arXiv2024

Low surface brightness galaxies from BASS+MzLS with Machine Learning

From $\sim$ 5000 deg$^{2}$ of the combination of the Beijing-Arizona Sky Survey (BASS) and Mayall $z$-band Legacy Survey (MzLS) which is also the northern sky region of the Dark Energy Spectroscopic Instrument (DESI) Legacy Imaging Surveys, we selected a sample of 31,825 candidates of low surface brightness galaxies (LSBGs) with the mean effective surface brightness 24.2 $< \barμ_{\rm eff,g} <$ 28.8 mag arcsec$^{\rm -2}$ and the half-light radius 2.5$^{\prime\prime}$ $< r_{\rm eff} <$ 20$^{\prime\prime}$ based on the released photometric catalogue and the machine learning model. The distribution of the LSBGs is of bimodality in the $g$ - $r$ color, indicating the two distinct populations of the blue ($g$ - $r <$ 0.60) and the red ($g$ - $r >$ 0.60) LSBGs. The blue LSBGs appear spiral, disk or irregular while the red LSBGs are spheroidal or ellipitcal and spatially clustered. This trend shows that the color has a strong correlation with galaxy morphology for LSBGs. In the spatial distribution, the blue LSBGs are more uniformly distributed while the red ones are highly clustered, indicating that red LSBGs preferentially populated denser environment than the blue LSBGs. Besides, both populations have consistent distribution of ellipticity (median $ε\sim$ 0.3), half-light radius (median $r_{\rm eff} \sim$ 4$^{\prime\prime}$), and Sersic index (median $n$ = 1), implying the dominance of the full sample by the round and disk galaxies. This sample has definitely extended the studies of LSBGs to a regime of lower surface brightness, fainter magnitude, and broader other properties than the previously SDSS-based samples.

astro-ph.GA↗

arXiv2024

An Enhanced Differential Grouping Method for Large-Scale Overlapping Problems

Large-scale overlapping problems are prevalent in practical engineering applications, and the optimization challenge is significantly amplified due to the existence of shared variables. Decomposition-based cooperative coevolution (CC) algorithms have demonstrated promising performance in addressing large-scale overlapping problems. However, current CC frameworks designed for overlapping problems rely on grouping methods for the identification of overlapping problem structures and the current grouping methods for large-scale overlapping problems fail to consider both accuracy and efficiency simultaneously. In this article, we propose a two-stage enhanced grouping method for large-scale overlapping problems, called OEDG, which achieves accurate grouping while significantly reducing computational resource consumption. In the first stage, OEDG employs a grouping method based on the finite differences principle to identify all subcomponents and shared variables. In the second stage, we propose two grouping refinement methods, called subcomponent union detection (SUD) and subcomponent detection (SD), to enhance and refine the grouping results. SUD examines the information of the subcomponents and shared variables obtained in the previous stage, and SD corrects inaccurate grouping results. To better verify the performance of the proposed OEDG, we propose a series of novel benchmarks that consider various properties of large-scale overlapping problems, including the topology structure, overlapping degree, and separability. Extensive experimental results demonstrate that OEDG is capable of accurately grouping different types of large-scale overlapping problems while consuming fewer computational resources. Finally, we empirically verify that the proposed OEDG can effectively improve the optimization performance of diverse large-scale overlapping problems.

cs.NE↗

arXiv2024

Gas-rich Ultra-diffuse Galaxies Are Originated from High Specific Angular Momentum

Ultra-diffuse galaxies, characterized by comparable effective radii to the Milky Way but possessing 100-1,000 times fewer stars, offer a unique opportunity to garner novel insights into the mechanisms governing galaxy formation. Nevertheless, the existing corpus of observational and simulation studies has not yet yielded a definitive constraint or comprehensive consensus on the formation mechanisms underlying ultra-diffuse galaxies. In this study, we delve into the properties of ultra-diffuse galaxies enriched with neutral hydrogen using a semi-analytic method, with the explicit aim of constraining existing ultra-diffuse galaxy formation models. We find that the gas-rich ultra-diffuse galaxies are statistically not failed $L^{\star}$ galaxies nor dark matter deficient galaxies. In statistical terms, these ultra-diffuse galaxies exhibit comparable halo concentration, but higher baryonic mass fraction, as well as higher stellar and gas specific angular momentum, in comparison to typical dwarf galaxy counterparts. Our analysis unveils that higher gas specific angular momentum serves as the underlying factor elucidating the observed heightened baryonic mass fractions, diminished star formation efficiency, expanded stellar disk sizes, and reduced stellar densities in ultra-diffuse galaxies. Our findings make significant contributions to advancing our knowledge of ultra-diffuse galaxy formation and shed light on the intricate interplay between gas dynamics and the evolution of galaxies.

astro-ph.GA↗

arXiv2024

Almost Optically Dark Galaxies in DECaLS (I): Detection, Optical Properties and Possible Origins

We report the discovery of eight optical counterparts of ALFALFA extragalactic objects from DECaLS, five of which are discovered for the first time. These objects were flagged as HI emission sources with no optical counterparts in SDSS before. Multi-band data reveal their unusual physical properties. They are faint and blue ($g-r=-0.35\sim0.55$), with quite low surface brightness ($μ_{\rm g,peak}=24.88\sim26.41\,{\rm mag}/{\rm arcsec}^2$), irregular morphologies, low stellar masses ($log_{10}(M_{*}/M_\odot)=5.27\sim7.15$), low star formation rates ($SFR=0.21\sim9.24\times10^{-3}\,{M_\odot}\,{\rm yr}^{-1}$), and remarkably high HI-to-stellar mass ratios ($log_{10}(M_{\rm HI}/M_{*}) = 1.72\sim3.22$, except AGC\,215415). They deviate from the scaling relations between HI and optical properties defined by the ALFALFA sample and the baryonic Tully-Fisher relation. They agree well with the main sequence of star-forming galaxies but exhibit low star-forming efficiency. Based on their physical properties and environments, we speculate that six of these objects may have originated from tidal processes, while the remaining two appear to have isolated origins. They may have had a relatively calm evolutionary history and only begun to form stars recently.

astro-ph.GA↗