arXiv subjects

Yu Zhang

Publications and source records attributed to Yu Zhang.

At least 649 records · Page 36Linked to original sources

arXiv2025

Signal Processing Meets SGD: From Momentum to Filter

In deep learning, stochastic gradient descent (SGD) and its momentum-based variants are widely used for optimization. However, the internal dynamics of these methods remain underexplored. In this paper, we analyze gradient behavior through a signal processing lens, isolating key factors that influence gradient updates and revealing a critical limitation: momentum techniques lack the flexibility to adequately balance bias and variance components in gradients, resulting in gradient estimation inaccuracies. To address this issue, we introduce a novel method SGDF (SGD with Filter) based on Wiener Filter principles, which derives an optimal time-varying gain to refine gradient updates by minimizing the mean square error in gradient estimation. This method yields an optimal first-order gradient estimate, effectively balancing noise reduction and signal preservation. Furthermore, our approach could extend to adaptive optimizers, enhancing their generalization potential. Empirical results show that SGDF achieves superior convergence and generalization compared to traditional momentum methods, and performs competitively with state-of-the-art optimizers.

cs.LG↗

arXiv2025

Protein Large Language Models: A Comprehensive Survey

Protein-specific large language models (Protein LLMs) are revolutionizing protein science by enabling more efficient protein structure prediction, function annotation, and design. While existing surveys focus on specific aspects or applications, this work provides the first comprehensive overview of Protein LLMs, covering their architectures, training datasets, evaluation metrics, and diverse applications. Through a systematic analysis of over 100 articles, we propose a structured taxonomy of state-of-the-art Protein LLMs, analyze how they leverage large-scale protein sequence data for improved accuracy, and explore their potential in advancing protein engineering and biomedical research. Additionally, we discuss key challenges and future directions, positioning Protein LLMs as essential tools for scientific discovery in protein science. Resources are maintained at https://github.com/Yijia-Xiao/Protein-LLM-Survey.

q-bio.BM↗

arXiv2025

BHViT: Binarized Hybrid Vision Transformer

Model binarization has made significant progress in enabling real-time and energy-efficient computation for convolutional neural networks (CNN), offering a potential solution to the deployment challenges faced by Vision Transformers (ViTs) on edge devices. However, due to the structural differences between CNN and Transformer architectures, simply applying binary CNN strategies to the ViT models will lead to a significant performance drop. To tackle this challenge, we propose BHViT, a binarization-friendly hybrid ViT architecture and its full binarization model with the guidance of three important observations. Initially, BHViT utilizes the local information interaction and hierarchical feature aggregation technique from coarse to fine levels to address redundant computations stemming from excessive tokens. Then, a novel module based on shift operations is proposed to enhance the performance of the binary Multilayer Perceptron (MLP) module without significantly increasing computational overhead. In addition, an innovative attention matrix binarization method based on quantization decomposition is proposed to evaluate the token's importance in the binarized attention matrix. Finally, we propose a regularization loss to address the inadequate optimization caused by the incompatibility between the weight oscillation in the binary layers and the Adam Optimizer. Extensive experimental results demonstrate that our proposed algorithm achieves SOTA performance among binary ViT methods.

cs.CV↗

arXiv2025

Knowledge-Decoupled Synergetic Learning: An MLLM based Collaborative Approach to Few-shot Multimodal Dialogue Intention Recognition

Few-shot multimodal dialogue intention recognition is a critical challenge in the e-commerce domainn. Previous methods have primarily enhanced model classification capabilities through post-training techniques. However, our analysis reveals that training for few-shot multimodal dialogue intention recognition involves two interconnected tasks, leading to a seesaw effect in multi-task learning. This phenomenon is attributed to knowledge interference stemming from the superposition of weight matrix updates during the training process. To address these challenges, we propose Knowledge-Decoupled Synergetic Learning (KDSL), which mitigates these issues by utilizing smaller models to transform knowledge into interpretable rules, while applying the post-training of larger models. By facilitating collaboration between the large and small multimodal large language models for prediction, our approach demonstrates significant improvements. Notably, we achieve outstanding results on two real Taobao datasets, with enhancements of 6.37\% and 6.28\% in online weighted F1 scores compared to the state-of-the-art method, thereby validating the efficacy of our framework.

cs.CL↗

arXiv2025

Discrete Differential Evolution Particle Swarm Optimization Algorithm for Energy Saving Flexible Job Shop Scheduling Problem Considering Machine Multi States

As the continuous deepening of low-carbon emission reduction policies, the manufacturing industries urgently need sensible energy-saving scheduling schemes to achieve the balance between improving production efficiency and reducing energy consumption. In energy-saving scheduling, reasonable machine states-switching is a key point to achieve expected goals, i.e., whether the machines need to switch speed between different operations, and whether the machines need to add extra setup time between different jobs. Regarding this matter, this work proposes a novel machine multi states-based energy saving flexible job scheduling problem (EFJSP-M), which simultaneously takes into account machine multi speeds and setup time. To address the proposed EFJSP-M, a kind of discrete differential evolution particle swarm optimization algorithm (D-DEPSO) is designed. In specific, D-DEPSO includes a hybrid initialization strategy to improve the initial population performance, an updating mechanism embedded with differential evolution operators to enhance population diversity, and a critical path variable neighborhood search strategy to expand the solution space. At last, based on datasets DPs and MKs, the experiment results compared with five state-of-the-art algorithms demonstrate the feasible of EFJSP-M and the superior of D-DEPSO.

cs.NE↗

arXiv2025

Radio Observation of the Pulsar Wind Nebula in SNR G11.2-0.3

Pulsar wind nebulae (PWNe) are important sources for understanding galactic high-energy processes, but it is controversial until now about how high-energy particles in PWNe are accelerated and transported. Lacking radio counterparts of X-ray PWNe (the proposed acceleration sites) introduce difficulties to better understandings in multi wavelengths. Our recent 3, 6, and 16\,cm high-resolution observations of G11.2$-$0.3 PWN with the Australia Telescope Compact Array (ATCA) uniquely show morphological similarity with its X-ray PWN (a torus/jet feature). Spectral indices of the radio torus and jet are around -0.09 and -0.10, respectively. Meanwhile for the jet region, the spectral break between radio and X-ray spectra implies particle acceleration mechanisms other than a diffusive shock acceleration. Polarization results suggest a helical B-field inside the jet, the equipartition B-field strength of which is below 100\,$μ$G.

astro-ph.HE↗

arXiv2025

RevTogether: Supporting Science Story Revision with Multiple AI Agents

As a popular form of science communication, science stories attract readers because they combine engaging narratives with comprehensible scientific knowledge. However, crafting such stories requires substantial skill and effort, as writers must navigate complex scientific concepts and transform them into coherent and accessible narratives tailored to audiences with varying levels of scientific literacy. To address the challenge, we propose RevTogether, a multi-agent system (MAS) designed to support revision of science stories with human-like AI agents (using GPT-4o). RevTogether allows AI agents to simulate affects in addition to providing comments and writing suggestions, while offering varying degrees of user agency. Our preliminary user study with non-expert writers (N=3) highlighted the need for transparency in AI agents' decision-making processes to support learning and suggested that emotional interactions could enhance human-AI collaboration in science storytelling.

cs.HC↗

arXiv2025

Higgs-strahlung at the LHC in the inert doublet model

$pp \to W^{\pm} h, Zh$ processes at the LHC are studied in the framework of the inert doublet model (IDM). To quantify the effects of the IDM and their observability in these processes we revisit the NLO (QCD and EW) predictions in the Standard Model (SM) and their uncertainty. Taking all available current constraints on the parameter space of the IDM, we consider both the case of the IDM providing a good Dark Matter (DM) candidate within the freeze-out mechanism as well as when the DM restrictions are relaxed. In the former, deviations from the SM of only a few per mil in these cross sections at the LHC are found and may not be measured. In the latter, the deviations can reach a few percents and should be observable. Smaller discrepancies in this case require that the theoretical uncertainties be improved, in particular those arising from the parton distribution functions (PDFs). We stress the importance of the photon-induced real corrections and the need for further improvement in the extraction of the photon PDF. The analysis also showcases the development and exploitation of our automated tool for the computation of one-loop electroweak and QCD corrections for a New Physics model with internal tests such as those concerning the soft and collinear parts provided through both dipole subtraction and phase space slicing besides tests for ultra-violet finiteness and gauge-parameter independence.

hep-ph↗

arXiv2025

Image Watermarks are Removable Using Controllable Regeneration from Clean Noise

Image watermark techniques provide an effective way to assert ownership, deter misuse, and trace content sources, which has become increasingly essential in the era of large generative models. A critical attribute of watermark techniques is their robustness against various manipulations. In this paper, we introduce a watermark removal approach capable of effectively nullifying state-of-the-art watermarking techniques. Our primary insight involves regenerating the watermarked image starting from a clean Gaussian noise via a controllable diffusion model, utilizing the extracted semantic and spatial features from the watermarked image. The semantic control adapter and the spatial control network are specifically trained to control the denoising process towards ensuring image quality and enhancing consistency between the cleaned image and the original watermarked image. To achieve a smooth trade-off between watermark removal performance and image consistency, we further propose an adjustable and controllable regeneration scheme. This scheme adds varying numbers of noise steps to the latent representation of the watermarked image, followed by a controlled denoising process starting from this noisy latent representation. As the number of noise steps increases, the latent representation progressively approaches clean Gaussian noise, facilitating the desired trade-off. We apply our watermark removal methods across various watermarking techniques, and the results demonstrate that our methods offer superior visual consistency/quality and enhanced watermark removal performance compared to existing regeneration approaches. Our code is available at https://github.com/yepengliu/CtrlRegen.

cs.CR↗

arXiv2025

Dynamical Diffusion: Learning Temporal Dynamics with Diffusion Models

Diffusion models have emerged as powerful generative frameworks by progressively adding noise to data through a forward process and then reversing this process to generate realistic samples. While these models have achieved strong performance across various tasks and modalities, their application to temporal predictive learning remains underexplored. Existing approaches treat predictive learning as a conditional generation problem, but often fail to fully exploit the temporal dynamics inherent in the data, leading to challenges in generating temporally coherent sequences. To address this, we introduce Dynamical Diffusion (DyDiff), a theoretically sound framework that incorporates temporally aware forward and reverse processes. Dynamical Diffusion explicitly models temporal transitions at each diffusion step, establishing dependencies on preceding states to better capture temporal dynamics. Through the reparameterization trick, Dynamical Diffusion achieves efficient training and inference similar to any standard diffusion model. Extensive experiments across scientific spatiotemporal forecasting, video prediction, and time series forecasting demonstrate that Dynamical Diffusion consistently improves performance in temporal predictive tasks, filling a crucial gap in existing methodologies. Code is available at this repository: https://github.com/thuml/dynamical-diffusion.

cs.LG↗

arXiv2025

Zero-Power Backscatter Sensing and Communication Proof-of-Concept

In this paper, we present an experimental setup to evaluate the performance of a radio frequency identification (RFID)-based integrated sensing and communication (ISAC) system. We focus on both the communication and sensing capabilities of the system. Our experiments evaluate the system's performance in various channel fading scenarios and with different substrate materials, including wood, plastic, wall, and glass. Additionally, we utilize radio tomographic imaging (RTI) to detect human motion by analyzing received signal strength indicator (RSSI) data. Our results demonstrate the impact of different materials and environments on RSSI and highlight the potential of RFID-based systems for effective sensing and communication in diverse applications.

eess.SP↗

arXiv2025

Net-Zero Integrated Sensing and Communication in Backscatter Systems

Future wireless networks targeted for improving spectral and energy efficiency, are expected to simultaneously provide sensing functionality and support low-power communications. This paper proposes a novel net-zero integrated sensing and communication (ISAC) model for backscatter systems, including an access point (AP), a net-zero device, and a user receiver. We fully utilize the backscatter mechanism for sensing and communication without additional power consumption and signal processing in the hardware device, which reduces the system complexity and makes it feasible for practical applications. To further optimize the system performance, we design a novel signal frame structure for the ISAC model that effectively mitigates communication interference at the transmitter, tag, and receiver. Additionally, we employ distributed antennas for sensing which can be placed flexibly to capture a wider range of signals from diverse angles and distances, thereby improving the accuracy of sensing. We derive theoretical expressions for the symbol error rate (SER) and tag's location detection probability, and provide a detailed analysis of how the system parameters, such as transmit power and tag's reflection coefficient, affect the system performance.

eess.SP↗

arXiv2025

Silicon photonic modulator circuit with programmable intensity and phase modulation response

Electro-optical modulators are essential components in optical communication systems. They encode an electrical waveform onto an optical carrier. However, their performance is often limited by inherent electro-optic processes and imperfections in existing integrated designs, which limit their adaptability to diverse applications. This paper presents a circuit-level programmable modulator design that addresses these challenges. The proposed modulator can generate both intensity and phase modulation, optimizing performance without altering the underlying design or constraining platform limitations. We explain and demonstrate the principle with both carrier depletion-based modulators and SiGe electro-absorption modulators on a silicon photonic platform. Experiments demonstrate precise control and optimization capabilities surpassing those of traditional modulator designs, marking a significant leap forward in adaptability and performance enhancement across intensity, phase, and modulation linearity, enabled by programmable photonics. In combination with on-chip monitors, our circuit can be self-calibrating. This programmable modulator circuit concept can be applied to modulators in different platforms, incorporating existing phase shifter designs, and act as a drop-in replacement in more complex circuits. It has the potential to be widely used in optical communication, LiDAR, microwave photonics, and other systems, meeting the increasing demands of various applications.

physics.optics↗

arXiv2025

ZuantuSet: A Collection of Historical Chinese Visualizations and Illustrations

Historical visualizations are a valuable resource for studying the history of visualization and inspecting the cultural context where they were created. When investigating historical visualizations, it is essential to consider contributions from different cultural frameworks to gain a comprehensive understanding. While there is extensive research on historical visualizations within the European cultural framework, this work shifts the focus to ancient China, a cultural context that remains underexplored by visualization researchers. To this aim, we propose a semi-automatic pipeline to collect, extract, and label historical Chinese visualizations. Through the pipeline, we curate ZuantuSet, a dataset with over 71K visualizations and 108K illustrations. We analyze distinctive design patterns of historical Chinese visualizations and their potential causes within the context of Chinese history and culture. We illustrate potential usage scenarios for this dataset, summarize the unique challenges and solutions associated with collecting historical Chinese visualizations, and outline future research directions.

cs.HC↗

arXiv2025

Physics-Informed Gradient Estimation for Accelerating Deep Learning based AC-OPF

The optimal power flow (OPF) problem can be rapidly and reliably solved by employing responsive online solvers based on neural networks. The dynamic nature of renewable energy generation and the variability of power grid conditions necessitate frequent neural network updates with new data instances. To address this need and reduce the time required for data preparation time, we propose a semi-supervised learning framework aided by data augmentation. In this context, ridge regression replaces the traditional solver, facilitating swift prediction of optimal solutions for the given input load demands. Additionally, to accelerate the backpropagation during training, we develop novel batch-mean gradient estimation approaches along with a reduced branch set to alleviate the complexity of gradient computation. Numerical simulations demonstrate that our neural network, equipped with the proposed gradient estimators, consistently achieves feasible and near-optimal solutions. These results underline the effectiveness of our approach for practical implementation in real-time OPF applications.

eess.SY↗

arXiv2025

ESANS: Effective and Semantic-Aware Negative Sampling for Large-Scale Retrieval Systems

Industrial recommendation systems typically involve a two-stage process: retrieval and ranking, which aims to match users with millions of items. In the retrieval stage, classic embedding-based retrieval (EBR) methods depend on effective negative sampling techniques to enhance both performance and efficiency. However, existing techniques often suffer from false negatives, high cost for ensuring sampling quality and semantic information deficiency. To address these limitations, we propose Effective and Semantic-Aware Negative Sampling (ESANS), which integrates two key components: Effective Dense Interpolation Strategy (EDIS) and Multimodal Semantic-Aware Clustering (MSAC). EDIS generates virtual samples within the low-dimensional embedding space to improve the diversity and density of the sampling distribution while minimizing computational costs. MSAC refines the negative sampling distribution by hierarchically clustering item representations based on multimodal information (visual, textual, behavioral), ensuring semantic consistency and reducing false negatives. Extensive offline and online experiments demonstrate the superior efficiency and performance of ESANS.

cs.IR↗

arXiv2025

The Graph Fractional Fourier Transform in Hilbert Space

Graph signal processing (GSP) leverages the inherent signal structure within graphs to extract high-dimensional data without relying on translation invariance. It has emerged as a crucial tool across multiple fields, including learning and processing of various networks, data analysis, and image processing. In this paper, we introduce the graph fractional Fourier transform in Hilbert space (HGFRFT), which provides additional fractional analysis tools for generalized GSP by extending Hilbert space and vertex domain Fourier analysis to fractional order. First, we establish that the proposed HGFRFT extends traditional GSP, accommodates graphs on continuous domains, and facilitates joint time-vertex domain transform while adhering to critical properties such as additivity, commutativity, and invertibility. Second, to process generalized graph signals in the fractional domain, we explore the theory behind filtering and sampling of signals in the fractional domain. Finally, our simulations and numerical experiments substantiate the advantages and enhancements yielded by the HGFRFT.

math.GM↗

arXiv2025

Towards a precision determination of the $X(6200)$ parameters from data

In a recent paper, Phys. Rev. Lett. {\bf 126}, 132001 (2021), the LHCb data on the di-$J/ψ$ production in proton-proton collisions were analysed in a coupled-channel framework based on double-vector-charmonium channels. This investigation identified a robust pole near the $J/ψJ/ψ$ threshold, tagged $X(6200)$, suggesting it as a new state. The present work extends that investigation by incorporating recent di-$J/ψ$ production data from the CMS and ATLAS Collaborations and performing a combined analysis of all three datasets. This study confirms the existence of the $X(6200)$, and its pole position is now determined with a higher precision than in the previous study, where only a single dataset was employed. The pole corresponding to the $X(6900)$ is also extracted, though its actual position and residue depend on a particular coupled-channel model employed and deviate from values reported in the experimental investigations. Moreover, we demonstrate that the currently available data do not allow one to determine whether there is an additional pole in the studied mass range. The di-electron width of the $X(6200)$ is estimated under different conjectures on its quantum numbers, and the number of such states that can be annually produced at the future Super $τ$-Charm Facility is estimated. The line shape in the complimentary $J/ψψ(2S)$ channel is discussed, and a good agreement with the ATLAS data is found.

hep-ph↗