arXiv subjects

Yi Xu

Publications and source records attributed to Yi Xu.

At least 199 records · Page 11Linked to original sources

arXiv2025

Studying Effects of Evolution on Charge Separation in Small Collision System with A MultiPhase Transport model

In relativistic high energy heavy-ion collisions, Chiral Magnetic Effect (CME) could produce a charge separation in QGP. The charge separation could survive into final hadron system during evolution, observed as correlator $Δγ$. This physics procedure could also occurs in small collision system if CME emerge in the QGP droplet. In this paper, we study the effects of QGP droplet evolution on charge separation in small collision system, based on AMPT model. Our calculation indicates that, with given initial charge separation, the effects of parton level evolution and hadron level evolution weaken the charge separation indeed, but there is still enough signals could remained. Furthermore, we found that, the contribution of background to $Δγ$ is negligible in small collision system. So, we propose the small collision system being a good place to confirm contribution of CME.

hep-ph↗

arXiv2025

PedDet: Adaptive Spectral Optimization for Multimodal Pedestrian Detection

Pedestrian detection in intelligent transportation systems has made significant progress but faces two critical challenges: (1) insufficient fusion of complementary information between visible and infrared spectra, particularly in complex scenarios, and (2) sensitivity to illumination changes, such as low-light or overexposed conditions, leading to degraded performance. To address these issues, we propose PedDet, an adaptive spectral optimization complementarity framework specifically enhanced and optimized for multispectral pedestrian detection. PedDet introduces the Multi-scale Spectral Feature Perception Module (MSFPM) to adaptively fuse visible and infrared features, enhancing robustness and flexibility in feature extraction. Additionally, the Illumination Robustness Feature Decoupling Module (IRFDM) improves detection stability under varying lighting by decoupling pedestrian and background features. We further design a contrastive alignment to enhance intermodal feature discrimination. Experiments on LLVIP and MSDS datasets demonstrate that PedDet achieves state-of-the-art performance, improving the mAP by 6.6% with superior detection accuracy even in low-light conditions, marking a significant step forward for road safety. Code will be available at https://github.com/AIGeeksGroup/PedDet.

cs.CV↗

arXiv2025

Enhanced Proton Acceleration via Petawatt Laguerre-Gaussian Lasers

High-energy, high-flux collimated proton beams with high repetition rates are critical for applications such as proton therapy, proton radiography, high-energy-density matter generation, and compact particle accelerators. However, achieving proton beam collimation has typically relied on complex and expensive target fabrication or precise control of auxiliary laser pulses, which poses significant limitations for high-repetition applications. Here, we demonstrate an all-optical method for collimated proton acceleration using a single femtosecond Laguerre-Gaussian (LG) laser with an intensity exceeding 1020 W/cm2 irradiating a simple planar target. Compared to conventional Gaussian laser-driven schemes, the maximum proton energy is enhanced by 60% (reaching 35 MeV) and beam divergence is much reduced. Particle-in-cell simulations reveal that a plasma jet is initially focused by the hollow electric sheath field of the LG laser, and then electrons in the jet are further collimated by self-generated magnetic fields. This process amplifies the charge-separation electric field between electrons and ions, leading to increased proton energy in the longitudinal direction and improved collimation in the transverse direction. This single-LG-laser-driven collimation mechanism offers a promising pathway for high-repetition, high-quality proton beam generation, with broad potential applications including proton therapy and fast ignition in inertial confinement fusion.

physics.plasm-ph↗

arXiv2025

HuRef: HUman-REadable Fingerprint for Large Language Models

Protecting the copyright of large language models (LLMs) has become crucial due to their resource-intensive training and accompanying carefully designed licenses. However, identifying the original base model of an LLM is challenging due to potential parameter alterations. In this study, we introduce HuRef, a human-readable fingerprint for LLMs that uniquely identifies the base model without interfering with training or exposing model parameters to the public. We first observe that the vector direction of LLM parameters remains stable after the model has converged during pretraining, with negligible perturbations through subsequent training steps, including continued pretraining, supervised fine-tuning, and RLHF, which makes it a sufficient condition to identify the base model. The necessity is validated by continuing to train an LLM with an extra term to drive away the model parameters' direction and the model becomes damaged. However, this direction is vulnerable to simple attacks like dimension permutation or matrix rotation, which significantly change it without affecting performance. To address this, leveraging the Transformer structure, we systematically analyze potential attacks and define three invariant terms that identify an LLM's base model. Due to the potential risk of information leakage, we cannot publish invariant terms directly. Instead, we map them to a Gaussian vector using an encoder, then convert it into a natural image using StyleGAN2, and finally publish the image. In our black-box setting, all fingerprinting steps are internally conducted by the LLMs owners. To ensure the published fingerprints are honestly generated, we introduced Zero-Knowledge Proof (ZKP). Experimental results across various LLMs demonstrate the effectiveness of our method. The code is available at https://github.com/LUMIA-Group/HuRef.

cs.CL↗

arXiv2025

Robust Fast Adaptation from Adversarially Explicit Task Distribution Generation

Meta-learning is a practical learning paradigm to transfer skills across tasks from a few examples. Nevertheless, the existence of task distribution shifts tends to weaken meta-learners' generalization capability, particularly when the training task distribution is naively hand-crafted or based on simple priors that fail to cover critical scenarios sufficiently. Here, we consider explicitly generative modeling task distributions placed over task identifiers and propose robustifying fast adaptation from adversarial training. Our approach, which can be interpreted as a model of a Stackelberg game, not only uncovers the task structure during problem-solving from an explicit generative model but also theoretically increases the adaptation robustness in worst cases. This work has practical implications, particularly in dealing with task distribution shifts in meta-learning, and contributes to theoretical insights in the field. Our method demonstrates its robustness in the presence of task subpopulation shifts and improved performance over SOTA baselines in extensive experiments. The code is available at the project site https://sites.google.com/view/ar-metalearn.

cs.LG↗

arXiv2024

SolidGS: Consolidating Gaussian Surfel Splatting for Sparse-View Surface Reconstruction

Gaussian splatting has achieved impressive improvements for both novel-view synthesis and surface reconstruction from multi-view images. However, current methods still struggle to reconstruct high-quality surfaces from only sparse view input images using Gaussian splatting. In this paper, we propose a novel method called SolidGS to address this problem. We observed that the reconstructed geometry can be severely inconsistent across multi-views, due to the property of Gaussian function in geometry rendering. This motivates us to consolidate all Gaussians by adopting a more solid kernel function, which effectively improves the surface reconstruction quality. With the additional help of geometrical regularization and monocular normal estimation, our method achieves superior performance on the sparse view surface reconstruction than all the Gaussian splatting methods and neural field methods on the widely used DTU, Tanks-and-Temples, and LLFF datasets.

cs.CV↗

arXiv2024

EMPRESS. X. Spatially resolved mass-metallicity relation in extremely metal-poor galaxies: evidence of episodic star-formation fueled by a metal-poor gas infall

Using the Subaru/FOCAS IFU capability, we examine the spatially resolved relationships between gas-phase metallicity, stellar mass, and star-formation rate surface densities (Sigma_* and Sigma_SFR, respectively) in extremely metal-poor galaxies (EMPGs) in the local universe. Our analysis includes 24 EMPGs, comprising 9,177 spaxels, which span a unique parameter space of local metallicity (12+log(O/H) = 6.9 to 7.9) and stellar mass surface density (Sigma_* ~ 10^5 to 10^7 Msun/kpc^2), extending beyond the range of existing large integral-field spectroscopic surveys. Through spatially resolved emission line diagnostics based on the [NII] BPT-diagram, we verify the absence of evolved active galactic nuclei in these EMPGs. Our findings reveal that, while the resolved mass-metallicity relation exhibits significant scatter in the low-mass regime, this scatter is closely correlated with local star-formation surface density. Specifically, metallicity decreases as Sigma_SFR increases for a given Sigma_*. Notably, half of the EMPGs show a distinct metal-poor horizontal branch on the resolved mass-metallicity relation. This feature typically appears at the peak clump with the highest Sigma_* and Sigma_SFR and is surrounded by a relatively metal-enriched ambient region. These findings support a scenario in which metal-poor gas infall fuels episodic star formation in EMPGs, consistent with the kinematic properties observed in these systems. In addition, we identify four EMPGs with exceptionally low central metallicities (12+log(O/H) <~ 7.2), which display only a metal-poor clump without a surrounding metal-rich region. This suggests that such ultra-low metallicity EMPGs, at less than a few percent of the solar metallicity, may serve as valuable analogs for galaxies in the early stages of galaxy evolution.

astro-ph.GA↗

arXiv2024

A new method to clarify contribution of chiral magnetic effect in small collision system $p^{\uparrow} + A$ involving a transversely polarized proton

In this paper, we propose a new experiment method to check contribution of chiral magnetic effect (CME). With experimental data of DIS involving transversely polarized proton, we have calculated the 3-D charge density inside the polarized proton, which is found to have a significant violation to spherical symmetry. Then we have calculated the property of electromagnetic field (E-M field) generated by a single transversely polarized proton ($p^{\uparrow}$). Based on them, the E-M field generated in small collision system $p^{\uparrow}+A$ are studied. We find that the orientation of this E-M field has a significant dependence on the polarization direction of the proton, and the correlator ($Δγ$ ) has also significant dependence on the angle between reaction plane and polarization direction. As background contribution are canceled comparing two collision geometry schemes, only contribution of CME is remained.

hep-ph↗

arXiv2024

JWST, ALMA, and Keck Spectroscopic Constraints on the UV Luminosity Functions at z~7-14: Clumpiness and Compactness of the Brightest Galaxies in the Early Universe

We present the number densities and physical properties of the bright galaxies spectroscopically confirmed at $z\sim7-14$. Our sample is composed of 60 galaxies at $z_\mathrm{spec}\sim7-14$, including recently-confirmed galaxies at $z_\mathrm{spec}=12.34-14.32$ with JWST, as well as new confirmations at $z_\mathrm{spec}=6.583-7.643$ with $-24< M_\mathrm{UV}< -21$ mag using ALMA and Keck. Our JWST/NIRSpec observations have also revealed that very bright galaxy candidates at $z\sim10-13$ identified from ground-based telescope images before JWST are passive galaxies at $z\sim3-4$, emphasizing the necessity of strict screening and spectroscopy in the selection of the brightest galaxies at $z>10$. The UV luminosity functions derived from these spectroscopic results are consistent with a double power-law function, showing tensions with theoretical models at the bright end. To understand the origin of the overabundance of bright galaxies, we investigate their morphologies using JWST/NIRCam high-resolution images obtained in various surveys including PRIMER and COSMOS-Web. We find that $\sim70\%$ of the bright galaxies at $z\sim7$ exhibit clumpy morphologies with multiple sub-components, suggesting merger-induced starburst activity, which is consistent with SED fitting results showing bursty star formation histories. At $z\gtrsim10$, bright galaxies are classified into two types of galaxies; extended ones with weak high-ionization emission lines, and compact ones with strong high-ionization lines including NIV]$λ$1486, indicating that at least two different processes (e.g., merger-induced starburst and compact star formation/AGN) are shaping the physical properties of the brightest galaxies at $z\gtrsim10$ and are responsible for their overabundance.

astro-ph.GA↗

arXiv2024

Relational Contrastive Learning and Masked Image Modeling for Scene Text Recognition

Context-aware methods have achieved remarkable advancements in supervised scene text recognition by leveraging semantic priors from words. Considering the heterogeneity of text and background in STR, we propose that such contextual priors can be reinterpreted as the relations between textual elements, serving as effective self-supervised labels for representation learning. However, textual relations are restricted to the finite size of the dataset due to lexical dependencies, which causes over-fitting problem, thus compromising the representation quality. To address this, our work introduces a unified framework of Relational Contrastive Learning and Masked Image Modeling for STR (RCMSTR), which explicitly models the enriched textual relations. For the RCL branch, we first introduce the relational rearrangement module to cultivate new relations on the fly. Based on this, we further conduct relational contrastive learning to model the intra- and inter-hierarchical relations for frames, sub-words and words. On the other hand, MIM can naturally boost the context information via masking, where we find that the block masking strategy is more effective for STR. For the effective integration of RCL and MIM, we also introduce a novel decoupling design aimed at mitigating the impact of masked images on contrastive learning. Additionally, to enhance the compatibility of MIM with CNNs, we propose the adoption of sparse convolutions and directly sharing the weights with dense convolutions in training. The proposed RCMSTR demonstrates superior performance in various evaluation protocols for different STR-related downstream tasks, outperforming the existing state-of-the-art self-supervised STR techniques. Ablation studies and qualitative experimental results further validate the effectiveness of our method. The code and pre-trained models will be available at https://github.com/ThunderVVV/RCMSTR .

cs.CV↗

arXiv2024

Pie: Pooling CPU Memory for LLM Inference

The rapid growth of LLMs has revolutionized natural language processing and AI analysis, but their increasing size and memory demands present significant challenges. A common solution is to spill over to CPU memory; however, traditional GPU-CPU memory swapping often results in higher latency and lower throughput. This paper introduces Pie, an LLM inference framework that addresses these challenges with performance-transparent swapping and adaptive expansion. By leveraging predictable memory access patterns and the high bandwidth of modern hardware like the NVIDIA GH200 Grace Hopper Superchip, Pie enables concurrent data swapping without affecting foreground computation, expanding effective memory without added latency. Adaptive expansion dynamically adjusts CPU memory allocation based on real-time information, optimizing memory usage and performance under varying conditions. Pie maintains low computation latency, high throughput, and high elasticity. Our experimental evaluation demonstrates that Pie achieves optimal swapping policy during cache warmup and effectively balances increased memory capacity with negligible impact on computation. With its extended capacity, Pie outperforms vLLM by up to 1.9X in throughput and 2X in latency. Additionally, Pie can reduce GPU memory usage by up to 1.67X while maintaining the same performance. Compared to FlexGen, an offline profiling-based swapping solution, Pie achieves magnitudes lower latency and 9.4X higher throughput.

cs.LG↗

arXiv2024

RIAV-MVS: Recurrent-Indexing an Asymmetric Volume for Multi-View Stereo

This paper presents a learning-based method for multi-view depth estimation from posed images. Our core idea is a "learning-to-optimize" paradigm that iteratively indexes a plane-sweeping cost volume and regresses the depth map via a convolutional Gated Recurrent Unit (GRU). Since the cost volume plays a paramount role in encoding the multi-view geometry, we aim to improve its construction both at pixel- and frame- levels. At the pixel level, we propose to break the symmetry of the Siamese network (which is typically used in MVS to extract image features) by introducing a transformer block to the reference image (but not to the source images). Such an asymmetric volume allows the network to extract global features from the reference image to predict its depth map. Given potential inaccuracies in the poses between reference and source images, we propose to incorporate a residual pose network to correct the relative poses. This essentially rectifies the cost volume at the frame level. We conduct extensive experiments on real-world MVS datasets and show that our method achieves state-of-the-art performance in terms of both within-dataset evaluation and cross-dataset generalization. Code available: https://github.com/oppo-us-research/riav-mvs.

cs.CV↗

arXiv2024

Cross-Fundus Transformer for Multi-modal Diabetic Retinopathy Grading with Cataract

Diabetic retinopathy (DR) is a leading cause of blindness worldwide and a common complication of diabetes. As two different imaging tools for DR grading, color fundus photography (CFP) and infrared fundus photography (IFP) are highly-correlated and complementary in clinical applications. To the best of our knowledge, this is the first study that explores a novel multi-modal deep learning framework to fuse the information from CFP and IFP towards more accurate DR grading. Specifically, we construct a dual-stream architecture Cross-Fundus Transformer (CFT) to fuse the ViT-based features of two fundus image modalities. In particular, a meticulously engineered Cross-Fundus Attention (CFA) module is introduced to capture the correspondence between CFP and IFP images. Moreover, we adopt both the single-modality and multi-modality supervisions to maximize the overall performance for DR grading. Extensive experiments on a clinical dataset consisting of 1,713 pairs of multi-modal fundus images demonstrate the superiority of our proposed method. Our code will be released for public access.

eess.IV↗

arXiv2024

Shopping MMLU: A Massive Multi-Task Online Shopping Benchmark for Large Language Models

Online shopping is a complex multi-task, few-shot learning problem with a wide and evolving range of entities, relations, and tasks. However, existing models and benchmarks are commonly tailored to specific tasks, falling short of capturing the full complexity of online shopping. Large Language Models (LLMs), with their multi-task and few-shot learning abilities, have the potential to profoundly transform online shopping by alleviating task-specific engineering efforts and by providing users with interactive conversations. Despite the potential, LLMs face unique challenges in online shopping, such as domain-specific concepts, implicit knowledge, and heterogeneous user behaviors. Motivated by the potential and challenges, we propose Shopping MMLU, a diverse multi-task online shopping benchmark derived from real-world Amazon data. Shopping MMLU consists of 57 tasks covering 4 major shopping skills: concept understanding, knowledge reasoning, user behavior alignment, and multi-linguality, and can thus comprehensively evaluate the abilities of LLMs as general shop assistants. With Shopping MMLU, we benchmark over 20 existing LLMs and uncover valuable insights about practices and prospects of building versatile LLM-based shop assistants. Shopping MMLU can be publicly accessed at https://github.com/KL4805/ShoppingMMLU. In addition, with Shopping MMLU, we host a competition in KDD Cup 2024 with over 500 participating teams. The winning solutions and the associated workshop can be accessed at our website https://amazon-kddcup24.github.io/.

cs.LG↗

arXiv2024

First Proof of Principle Experiment for Muon Production with Ultrashort High Intensity Laser

Muons, which play a crucial role in both fundamental and applied physics, have traditionally been generated through proton accelerators or from cosmic rays. With the advent of ultra-short high-intensity lasers capable of accelerating electrons to GeV levels, it has become possible to generate muons in laser laboratories. In this work, we show the first proof of principle experiment for novel muon production with an ultra-short, high-intensity laser device through GeV electron beam bombardment on a lead converter target. The muon physical signal is confirmed by measuring its lifetime which is the first clear demonstration of laser-produced muons. Geant4 simulations were employed to investigate the photo-production, electro-production, and Bethe-Heitler processes response for muon generation and their subsequent detection. The results show that the dominant contributions of muons are attributed to the photo-production/electro-production and a significant yield of muons up to 0.01 $μ$/$e^-$ out of the converter target could be achieved. This laser muon source features compact, ultra-short pulse and high flux. Moreover, its implementation in a small laser laboratory is relatively straightforward, significantly reducing the barriers to entry for research in areas such as muonic X-ray elemental analysis, muon spin spectroscopy and so on.

physics.acc-ph↗

arXiv2024

RepEval: Effective Text Evaluation with LLM Representation

The era of Large Language Models (LLMs) raises new demands for automatic evaluation metrics, which should be adaptable to various application scenarios while maintaining low cost and effectiveness. Traditional metrics for automatic text evaluation are often tailored to specific scenarios, while LLM-based evaluation metrics are costly, requiring fine-tuning or rely heavily on the generation capabilities of LLMs. Besides, previous LLM-based metrics ignore the fact that, within the space of LLM representations, there exist direction vectors that indicate the estimation of text quality. To this end, we introduce RepEval, a metric that leverages the projection of LLM representations for evaluation. Through simple prompt modifications, RepEval can easily transition to various tasks, requiring only minimal sample pairs for direction vector construction. Results on fourteen datasets across two evaluation tasks demonstrate the high effectiveness of our method, which exhibits a higher correlation with human judgments than previous methods, even in complex evaluation scenarios involving pair-wise selection under nuanced aspects. Our work underscores the richness of information regarding text quality embedded within LLM representations, offering insights for the development of new metrics.

cs.CL↗

arXiv2024

Evolutionary Contrastive Distillation for Language Model Alignment

The ability of large language models (LLMs) to execute complex instructions is essential for their real-world applications. However, several recent studies indicate that LLMs struggle with challenging instructions. In this paper, we propose Evolutionary Contrastive Distillation (ECD), a novel method for generating high-quality synthetic preference data designed to enhance the complex instruction-following capability of language models. ECD generates data that specifically illustrates the difference between a response that successfully follows a set of complex instructions and a response that is high-quality, but nevertheless makes some subtle mistakes. This is done by prompting LLMs to progressively evolve simple instructions to more complex instructions. When the complexity of an instruction is increased, the original successful response to the original instruction becomes a "hard negative" response for the new instruction, mostly meeting requirements of the new instruction, but barely missing one or two. By pairing a good response with such a hard negative response, and employing contrastive learning algorithms such as DPO, we improve language models' ability to follow complex instructions. Empirically, we observe that our method yields a 7B model that exceeds the complex instruction-following performance of current SOTA 7B models and is competitive even with open-source 70B models.

cs.LG↗

arXiv2024

Dynamics of a Galaxy at z > 10 Explored by JWST Integral Field Spectroscopy: Hints of Rotating Disk Suggesting Weak Feedback

We investigate the dynamics of GN-z11, a luminous galaxy at $z=10.60$, carefully analyzing the public deep integral field spectroscopy (IFS) data taken with JWST NIRSpec IFU. While the observations of the IFS data originally targeted a He II clump near GN-z11, we find that CIII]$λλ$1907,1909 emission from ionized gas at GN-z11 is bright and spatially extended significantly beyond the point-spread function (PSF). The spatially extended CIII emission of GN-z11 shows a velocity gradient, red- and blue-shifted components in the north and south directions, respectively, which cannot be explained by the variation of [CIII]$λ$1907/CIII]$λ$1909 line ratios. Assuming the velocity gradient is produced by disk rotation, we perform forward modeling with GalPak$^{3D}$, including the effects of PSF smearing and line blending, and obtain a rotation velocity of $v_{rot}=257^{+138}_{-117}$ km s$^{-1}$, a velocity dispersion of $σ_v=91^{+18}_{-32}$ km s$^{-1}$, and a ratio of $v_{rot}/σ_v=2.83^{+1.82}_{-1.41}$. The $v_{rot}/σ_v$ value would suggest a rotation-dominated disk existing at $z>10$ albeit with the large uncertainties. The rotation velocity agrees with those of numerical simulations predicting a rotating disk formed in the early universe under the condition of mass compaction and weak feedback. While the velocity gradient is consistent with the rotating disk solution, we recognize that galactic outflows can also explain the velocity gradient as well as the extended morphology and the high velocity dispersion found in the outskirt. Higher S/N and resolution data are necessary to conclude the physical origin of the velocity gradient in GN-z11.

astro-ph.GA↗