arXiv subjects

Qi Guo

Publications and source records attributed to Qi Guo.

At least 145 records · Page 8Linked to original sources

arXiv2025

Probing Nuclear Excitation by Electron Capture in an Electron Beam Ion Trap with Non-destructive Isomer Detection via Precision Mass Spectrometry

Nuclear excitation by electron capture (NEEC) is an important nuclear excitation mechanism which still lacks conclusive experimental verification. This is primarily attributed to strong background x-/$γ$-ray noise and competing nuclear excitation processes which would overshadow the signals in various environments that NEEC takes place. Here, we propose an experimental approach to observe the NEEC process within a background-free environment. Through collisions with a highly-compressed mono-energetic electron beam in an electron beam ion trap, nuclei may get excited to a long-lived isomeric state via the NEEC process. Subsequently, ions can be extracted and Penning-trap mass spectrometry employed to unambiguously detect the isomer. Our study focuses on the promising candidate $^{189}\mathrm{Os}$, demonstrating measurable detection rates of the NEEC process and discussing the feasibility of the proposed approach. This new approach for observing the NEEC process may be realized in the near future.

nucl-ex↗

arXiv2024

InverseCoder: Self-improving Instruction-Tuned Code LLMs with Inverse-Instruct

Recent advancements in open-source code large language models (LLMs) have been driven by fine-tuning on the data generated from powerful closed-source LLMs, which are expensive to obtain. This paper explores whether it is possible to use a fine-tuned open-source model to generate additional data to augment its instruction-tuning dataset. We make two observations: (1) A code snippet can serve as the response to different instructions. (2) Instruction-tuned code LLMs perform better at translating code into instructions than the reverse. Based on these observations, we propose Inverse-Instruct, a data augmentation technique that uses a fine-tuned LLM to generate additional instructions of code responses from its own training dataset. The additional instruction-response pairs are added to the original dataset, and a stronger code LLM can be obtained by fine-tuning on the augmented dataset. We empirically validate Inverse-Instruct on a range of open-source code models (e.g. CodeLlama-Python and DeepSeek-Coder) and benchmarks (e.g., HumanEval(+), MBPP(+), DS-1000 and MultiPL-E), showing it consistently improves the base models.

cs.CL↗

arXiv2024

Efficient Generation of Targeted and Transferable Adversarial Examples for Vision-Language Models Via Diffusion Models

Adversarial attacks, particularly \textbf{targeted} transfer-based attacks, can be used to assess the adversarial robustness of large visual-language models (VLMs), allowing for a more thorough examination of potential security flaws before deployment. However, previous transfer-based adversarial attacks incur high costs due to high iteration counts and complex method structure. Furthermore, due to the unnaturalness of adversarial semantics, the generated adversarial examples have low transferability. These issues limit the utility of existing methods for assessing robustness. To address these issues, we propose AdvDiffVLM, which uses diffusion models to generate natural, unrestricted and targeted adversarial examples via score matching. Specifically, AdvDiffVLM uses Adaptive Ensemble Gradient Estimation to modify the score during the diffusion model's reverse generation process, ensuring that the produced adversarial examples have natural adversarial targeted semantics, which improves their transferability. Simultaneously, to improve the quality of adversarial examples, we use the GradCAM-guided Mask method to disperse adversarial semantics throughout the image rather than concentrating them in a single area. Finally, AdvDiffVLM embeds more target semantics into adversarial examples after multiple iterations. Experimental results show that our method generates adversarial examples 5x to 10x faster than state-of-the-art transfer-based adversarial attacks while maintaining higher quality adversarial examples. Furthermore, compared to previous transfer-based adversarial attacks, the adversarial examples generated by our method have better transferability. Notably, AdvDiffVLM can successfully attack a variety of commercial VLMs in a black-box environment, including GPT-4V.

cs.CV↗

arXiv2024

Ultra Diffuse Dwarf Galaxies Hosting Pseudo-bulges

By analyzing data from DESI Legacy Imaging Survey of the dwarf galaxies in the Arecibo Legacy Fast Alfa Survey, we have identified five ultra-diffuse galaxies (UDGs) featuring central pseudo-bulges. These UDGs display blue pseudo-bulges with Sérsic indices $n<2.5$ and effective radii spanning 300-700 pc, along with bluer thin stellar disks exhibiting low surface brightness and expansive effective radii that align with the UDG definition. The rotation velocities of these UDGs, determined using HI line widths and optical inclinations, exceed those of most dwarf galaxies of similar mass, suggesting the high halo spins or substantial dark matter halos. We propose that these UDGs likely formed through mergers of dwarf galaxies lacking old stars in their progenitors, resulting in the development of central bulge-like structures during starbursts triggered by the mergers, while also enhancing their halo spin. Subsequent gas accretion facilitated the formation of extended stellar disks. It is also worth noting the possibility that these UDGs could alternatively represent ``failed $L^{\star}$ galaxies'' with massive dark matter halos but reduced star formation efficiencies. If future high-resolution HI observations confirm the presence of massive halos around these UDGs, they may have formed due to intense AGN feedback in the early universe, and may be the descendants of ``little red dots'' observed by the James Webb Space Telescope, which are characterized by heightened central black hole masses and intensified accretion and feedback processes in the early universe.

astro-ph.GA↗

arXiv2024

The density profile of Milky Way dark matter halo constrained from the OGLE microlensing sky map

We report the detection of a 282 $^{+34}_{-31}$ pc-sized core in the center of Milky Way dark matter halo at $68\%$ confidence level by using the micro-lensing event rate sky map data from the Optical Gravitational Lensing Experiment (OGLE) survey. We apply the spacial information of the microlensing sky map and model it with the detailed Milky Way dark matter halo Core/Cusp profile, and the fraction of dark matter in the form of Mini Dark Matter Structure (MDMS, $f_{\rm MDMS}=Ω_{\rm MDMS}/Ω_{\rm DM}$), e.g. primordial black hole, earth-mass subhalos, floating planets and so on. We find that this sky map can constrain both $f_{\rm MDMS}$ and the core size simultaneously without strong degeneracy while fully considering mass function of Milky Way stellar components from both the bulge and disk.

astro-ph.GA↗

arXiv2024

InputSnatch: Stealing Input in LLM Services via Timing Side-Channel Attacks

Large language models (LLMs) possess extensive knowledge and question-answering capabilities, having been widely deployed in privacy-sensitive domains like finance and medical consultation. During LLM inferences, cache-sharing methods are commonly employed to enhance efficiency by reusing cached states or responses for the same or similar inference requests. However, we identify that these cache mechanisms pose a risk of private input leakage, as the caching can result in observable variations in response times, making them a strong candidate for a timing-based attack hint. In this study, we propose a novel timing-based side-channel attack to execute input theft in LLMs inference. The cache-based attack faces the challenge of constructing candidate inputs in a large search space to hit and steal cached user queries. To address these challenges, we propose two primary components. The input constructor employs machine learning techniques and LLM-based approaches for vocabulary correlation learning while implementing optimized search mechanisms for generalized input construction. The time analyzer implements statistical time fitting with outlier elimination to identify cache hit patterns, continuously providing feedback to refine the constructor's search strategy. We conduct experiments across two cache mechanisms and the results demonstrate that our approach consistently attains high attack success rates in various applications. Our work highlights the security vulnerabilities associated with performance optimizations, underscoring the necessity of prioritizing privacy and security alongside enhancements in LLM inference.

cs.CR↗

arXiv2024

Static Microlensing: Concept, Method and Candidates

We propose a novel microlensing event search method that differs from either the traditional time domain method, astrometric microlensing, or the parallax microlensing method. Our method assumes that stars with nearly identical "genes" - normalized Spectral Energy Distributions (SED) bear the same luminosity within the intrinsic scatter due to stellar properties. Given a sample of stars with similar normalized SEDs, the outliers in luminosity distribution can be considered microlensing events by excluding other possible variations. In this case, we can select microlensing events from archive data rather than time domain monitoring the sky, which we describe as static microlensing. Following this concept, we collect the data from Gaia DR3 and SDSS DR16 from the northern galactic cap at high galactic latitudes. This area is not preferable for normal microlensing search due to the low stellar density and, therefore, low discovery rate. By applying a similarity search algorithm, we find 5 microlensing candidates in the Galactic halo.

astro-ph.GA↗

arXiv2024

Cosmological Prediction of the Void and Galaxy Clustering Measurements in the CSST Spectroscopic Survey

The void power spectrum is related to the clustering of low-density regions in the large-scale structure (LSS) of the Universe, and can be used as an effective cosmological probe to extract the information of the LSS. We generate the galaxy mock catalogs from Jiutian simulation, and identify voids using the watershed algorithm for studying the cosmological constraint strength of the China Space Station Telescope (CSST) spectroscopic survey. The galaxy and void auto power spectra and void-galaxy cross power spectra at $z=0.3$, 0.6, and 0.9 are derived from the mock catalogs. To fit the full power spectra, we propose to use the void average effective radius at a given redshift to simplify the theoretical model, and adopt the Markov Chain Monte Carlo (MCMC) technique to implement the constraints on the cosmological and void parameters. The systematical parameters, such as galaxy and void biases, and noise terms in the power spectra are also included in the fitting process. We find that our theoretical model can correctly extract the cosmological information from the galaxy and void power spectra, which demonstrates its feasibility and effectivity. The joint constraint accuracy of the cosmological parameters can be improved by $\sim20\%$ compared to that from the galaxy power spectrum only. The fitting results of the void density profile and systematical parameters are also well constrained and consistent with the expectation. This indicates that the void clustering measurement can be an effective complement to the galaxy clustering probe, especially for the next generation galaxy surveys.

astro-ph.CO↗

arXiv2024

Generating entanglement of two acoustic modes by driving the qubit in circuit quantum acoustodynamics system

We propose how to generate the entanglement of two long-lived phonon modes in a circuit quantum acoustodynamics system, which consists of a multi-mode high-frequency bulk acoustic wave resonator and a transmon-type superconducting qubit. Two acoustic modes couple to the qubit through piezoelectric interaction, and the qubit is driven by a microwave field. Under the condition of far detuning between the qubit and acoustic modes, the qubit can be eliminated adiabatically, and thus establishing the indirect interaction between the two acoustic modes. We demonstrate that such the indirect interaction can be the parametric-amplification-type interaction by appropriately choosing the drive frequency and strength, so the entanglement between acoustic modes can be created by the direct unitary evolution. We numerically analyze the parameter conditions for generating the entanglement in detail and evaluate the influence of system dissipations and noise. The results show that the scheme can be realized using currently available parameters and has strong robustness against the dissipations and environmental temperature. This work may provide efficient resource for the quantum information processing based on the phononic systems.

quant-ph↗

arXiv2024

Apostle--Auriga: Effects of stellar feedback subgrid models on the evolution of angular momentum in disc galaxies

Utilizing the Apostle--Auriga simulations, which start from the same zoom-in initial conditions of Local Group-like systems but run with different galaxy formation subgrid models and hydrodynamic solvers, we study the impact of stellar feedback models on the evolution of angular momentum in disc galaxies. At $z = 0$, Auriga disc galaxies tend to exhibit higher specific angular momenta compared to their cross-matched Apostle counterparts. By tracing the evolution history of the Lagrangian mass tracers of the in-situ star particles in the $z = 0$ galaxies, we find that the specific angular momentum distributions of the gas tracers from the two simulations at the halo accretion time are relatively similar. The present-day angular momentum difference is mainly driven by the physical processes occurring inside dark matter haloes, especially galactic fountains. Due to the different subgrid implementations of stellar feedback processes, Auriga galaxies contain a high fraction of gas that has gone through recycled fountain (${\sim} 65$ per cent) which could acquire angular momentum through mixing with the high angular momentum circumgalactic medium (CGM). In Apostle, however, the fraction of gas that has undergone the recycled fountain process is significantly lower (down to ${\sim} 20$ per cent for Milky Way-sized galaxies) and the angular momentum acquisition from the CGM is marginal. As a result, the present-day Auriga galaxies overall have higher specific angular momenta.

astro-ph.GA↗

arXiv2024

Forgetting Through Transforming: Enabling Federated Unlearning via Class-Aware Representation Transformation

Federated Unlearning (FU) enables clients to selectively remove the influence of specific data from a trained federated learning model, addressing privacy concerns and regulatory requirements. However, existing FU methods often struggle to balance effective erasure with model utility preservation, especially for class-level unlearning in non-IID settings. We propose Federated Unlearning via Class-aware Representation Transformation (FUCRT), a novel method that achieves unlearning through class-aware representation transformation. FUCRT employs two key components: (1) a transformation class selection strategy to identify optimal forgetting directions, and (2) a transformation alignment technique using dual class-aware contrastive learning to ensure consistent transformations across clients. Extensive experiments on four datasets demonstrate FUCRT's superior performance in terms of erasure guarantee, model utility preservation, and efficiency. FUCRT achieves complete (100\%) erasure of unlearning classes while maintaining or improving performance on remaining classes, outperforming state-of-the-art baselines across both IID and Non-IID settings. Analysis of the representation space reveals FUCRT's ability to effectively merge unlearning class representations with the transformation class from remaining classes, closely mimicking the model retrained from scratch.

cs.LG↗

arXiv2024

An Empirical Study of Training ID-Agnostic Multi-modal Sequential Recommenders

Sequential Recommendation (SR) aims to predict future user-item interactions based on historical interactions. While many SR approaches concentrate on user IDs and item IDs, the human perception of the world through multi-modal signals, like text and images, has inspired researchers to delve into constructing SR from multi-modal information without using IDs. However, the complexity of multi-modal learning manifests in diverse feature extractors, fusion methods, and pre-trained models. Consequently, designing a simple and universal \textbf{M}ulti-\textbf{M}odal \textbf{S}equential \textbf{R}ecommendation (\textbf{MMSR}) framework remains a formidable challenge. We systematically summarize the existing multi-modal related SR methods and distill the essence into four core components: visual encoder, text encoder, multimodal fusion module, and sequential architecture. Along these dimensions, we dissect the model designs, and answer the following sub-questions: First, we explore how to construct MMSR from scratch, ensuring its performance either on par with or exceeds existing SR methods without complex techniques. Second, we examine if MMSR can benefit from existing multi-modal pre-training paradigms. Third, we assess MMSR's capability in tackling common challenges like cold start and domain transferring. Our experiment results across four real-world recommendation scenarios demonstrate the great potential ID-agnostic multi-modal sequential recommendation. Our framework can be found at: https://github.com/MMSR23/MMSR.

cs.IR↗

arXiv2024

Cambricon-LLM: A Chiplet-Based Hybrid Architecture for On-Device Inference of 70B LLM

Deploying advanced large language models on edge devices, such as smartphones and robotics, is a growing trend that enhances user data privacy and network connectivity resilience while preserving intelligent capabilities. However, such a task exhibits single-batch computing with incredibly low arithmetic intensity, which poses the significant challenges of huge memory footprint and bandwidth demands on limited edge resources. To address these issues, we introduce Cambricon-LLM, a chiplet-based hybrid architecture with NPU and a dedicated NAND flash chip to enable efficient on-device inference of 70B LLMs. Such a hybrid architecture utilizes both the high computing capability of NPU and the data capacity of the NAND flash chip, with the proposed hardware-tiling strategy that minimizes the data movement overhead between NPU and NAND flash chip. Specifically, the NAND flash chip, enhanced by our innovative in-flash computing and on-die ECC techniques, excels at performing precise lightweight on-die processing. Simultaneously, the NPU collaborates with the flash chip for matrix operations and handles special function computations beyond the flash's on-die processing capabilities. Overall, Cambricon-LLM enables the on-device inference of 70B LLMs at a speed of 3.44 token/s, and 7B LLMs at a speed of 36.34 token/s, which is over 22X to 45X faster than existing flash-offloading technologies, showing the potentiality of deploying powerful LLMs in edge devices.

cs.AR↗

arXiv2024

Depth from Coupled Optical Differentiation

We propose depth from coupled optical differentiation, a low-computation passive-lighting 3D sensing mechanism. It is based on our discovery that per-pixel object distance can be rigorously determined by a coupled pair of optical derivatives of a defocused image using a simple, closed-form relationship. Unlike previous depth-from-defocus (DfD) methods that leverage spatial derivatives of the image to estimate scene depths, the proposed mechanism's use of only optical derivatives makes it significantly more robust to noise. Furthermore, unlike many previous DfD algorithms with requirements on aperture code, this relationship is proved to be universal to a broad range of aperture codes. We build the first 3D sensor based on depth from coupled optical differentiation. Its optical assembly includes a deformable lens and a motorized iris, which enables dynamic adjustments to the optical power and aperture radius. The sensor captures two pairs of images: one pair with a differential change of optical power and the other with a differential change of aperture scale. From the four images, a depth and confidence map can be generated with only 36 floating point operations per output pixel (FLOPOP), more than ten times lower than the previous lowest passive-lighting depth sensing solution to our knowledge. Additionally, the depth map generated by the proposed sensor demonstrates more than twice the working range of previous DfD methods while using significantly lower computation.

cs.CV↗

arXiv2024

Void Number Counts as a Cosmological Probe for the Large-Scale Structure

Void number counts (VNC) indicates the number of low-density regions in the large-scale structure (LSS) of the Universe, and we propose to use it as an effective cosmological probe. By generating the galaxy mock catalog based on Jiutian simulations and considering the spectroscopic survey strategy and instrumental design of the China Space Station Telescope (CSST), which can reach a magnitude limit $\sim$23 AB mag and spectral resolution $R\gtrsim200$ with a sky coverage 17,500 deg$^2$, we identify voids using the watershed algorithm without any assumption of void shape, and obtain the mock void catalog and data of the VNC in six redshift bins from $z=0.3$ to1.3. We use the Markov Chain Monte Carlo (MCMC) method to constrain the cosmological and VNC parameters. The void linear underdensity threshold $δ_{\rm v}$ in the theoretical model is set to be a free parameter at a given redshift to fit the VNC data and explore its redshift evolution. We find that, the VNC can correctly derive the cosmological information, and the constraint strength on the cosmological parameters is comparable to that from the void size function (VSF) method, which can reach a few percentage levels in the CSST full spectroscopic survey. This is because that, since the VNC is not sensitive to void shape, the modified theoretical model can match the data better by integrating over void features, and more voids could be included in the VNC analysis by applying simpler selection criteria, which will improve the statistical significance. It indicates that the VNC can be an effective cosmological probe for exploring the LSS.

astro-ph.CO↗

arXiv2024

Understanding and Modeling Job Marketplace with Pretrained Language Models

Job marketplace is a heterogeneous graph composed of interactions among members (job-seekers), companies, and jobs. Understanding and modeling job marketplace can benefit both job seekers and employers, ultimately contributing to the greater good of the society. However, existing graph neural network (GNN)-based methods have shallow understandings of the associated textual features and heterogeneous relations. To address the above challenges, we propose PLM4Job, a job marketplace foundation model that tightly couples pretrained language models (PLM) with job market graph, aiming to fully utilize the pretrained knowledge and reasoning ability to model member/job textual features as well as various member-job relations simultaneously. In the pretraining phase, we propose a heterogeneous ego-graph-based prompting strategy to model and aggregate member/job textual features based on the topological structure around the target member/job node, where entity type embeddings and graph positional embeddings are introduced accordingly to model different entities and their heterogeneous relations. Meanwhile, a proximity-aware attention alignment strategy is designed to dynamically adjust the attention of the PLM on ego-graph node tokens in the prompt, such that the attention can be better aligned with job marketplace semantics. Extensive experiments at LinkedIn demonstrate the effectiveness of PLM4Job.

cs.IR↗

arXiv2024

Differentially Private Methods for Compositional Data

Confidential data, such as electronic health records, activity data from wearable devices, and geolocation data, are becoming increasingly prevalent. Differential privacy provides a framework to conduct statistical analyses while mitigating the risk of leaking private information. Compositional data, which consist of vectors with positive components that add up to a constant, have received little attention in the differential privacy literature. This article proposes differentially private approaches for analyzing compositional data using the Dirichlet distribution. We explore several methods, including Bayesian and bootstrap procedures. For the Bayesian methods, we consider posterior inference techniques based on Markov Chain Monte Carlo, Approximate Bayesian Computation, and asymptotic approximations. We conduct an extensive simulation study to compare these approaches and make evidence-based recommendations. Finally, we apply the methodology to a data set from the American Time Use Survey.

stat.ME↗

arXiv2024

TensorTEE: Unifying Heterogeneous TEE Granularity for Efficient Secure Collaborative Tensor Computing

Heterogeneous collaborative computing with NPU and CPU has received widespread attention due to its substantial performance benefits. To ensure data confidentiality and integrity during computing, Trusted Execution Environments (TEE) is considered a promising solution because of its comparatively lower overhead. However, existing heterogeneous TEE designs are inefficient for collaborative computing due to fine and different memory granularities between CPU and NPU. 1) The cacheline granularity of CPU TEE intensifies memory pressure due to its extra memory access, and 2) the cacheline granularity MAC of NPU escalates the pressure on the limited memory storage. 3) Data transfer across heterogeneous enclaves relies on the transit of non-secure regions, resulting in cumbersome re-encryption and scheduling. To address these issues, we propose TensorTEE, a unified tensor-granularity heterogeneous TEE for efficient secure collaborative tensor computing. First, we virtually support tensor granularity in CPU TEE to eliminate the off-chip metadata access by detecting and maintaining tensor structures on-chip. Second, we propose tensor-granularity MAC management with predictive execution to avoid computational stalls while eliminating off-chip MAC storage and access. Moreover, based on the unified granularity, we enable direct data transfer without re-encryption and scheduling dilemmas. Our evaluation is built on enhanced Gem5 and a cycle-accurate NPU simulator. The results show that TensorTEE improves the performance of Large Language Model (LLM) training workloads by 4.0x compared to existing work and incurs only 2.1% overhead compared to non-secure training, offering a practical security assurance for LLM training.

cs.CR↗