arXiv subjects

He Li

Publications and source records attributed to He Li.

At least 73 records · Page 4Linked to original sources

arXiv2025

Physics-informed deep learning for infectious disease forecasting

Accurate forecasting of contagious diseases is critical for public health policymaking and pandemic preparedness. We propose a new infectious disease forecasting model based on physics-informed neural networks (PINNs), an emerging scientific machine learning approach. By embedding a compartmental model into the loss function, our method integrates epidemiological theory with data, helping to prevent model overfitting. We further enhance the model with a sub-network that accounts for covariates such as mobility and cumulative vaccine doses, which influence the transmission rate. Using state-level COVID-19 data from California, we demonstrate that the PINN model accurately predicts cases, deaths, and hospitalizations, aligning well with existing benchmarks. Notably, the PINN model outperforms naive baseline forecasts and several sequence deep learning models, including Recurrent Neural Networks (RNNs), Long Short-Term Memory (LSTM) networks, Gated Recurrent Units (GRUs), and Transformers. It also achieves performance comparable to a sophisticated Gaussian infection state forecasting model that combines compartmental dynamics, a data observation model, and parameter regression. However, the PINN model features a simpler structure and is easier to implement. In summary, we systematically evaluate the PINN model's ability to forecast infectious disease dynamics, demonstrating its potential as an efficient computational tool to strengthen forecasting capabilities.

cs.LG↗

arXiv2025

Green Robotic Mixed Reality with Gaussian Splatting

Realizing green communication in robotic mixed reality (RoboMR) systems presents a challenge, due to the necessity of uploading high-resolution images at high frequencies through wireless channels. This paper proposes Gaussian splatting (GS) RoboMR (GSRMR), which achieves a lower energy consumption and makes a concrete step towards green RoboMR. The crux to GSRMR is to build a GS model which enables the simulator to opportunistically render a photo-realistic view from the robot's pose, thereby reducing the need for excessive image uploads. Since the GS model may involve discrepancies compared to the actual environments, a GS cross-layer optimization (GSCLO) framework is further proposed, which jointly optimizes content switching (i.e., deciding whether to upload image or not) and power allocation across different frames. The GSCLO problem is solved by an accelerated penalty optimization (APO) algorithm. Experiments demonstrate that the proposed GSRMR reduces the communication energy by over 10x compared with RoboMR. Furthermore, the proposed GSRMR with APO outperforms extensive baseline schemes, in terms of peak signal-to-noise ratio (PSNR) and structural similarity index measure (SSIM).

cs.RO↗

arXiv2025

SparseDM: Toward Sparse Efficient Diffusion Models

Diffusion models represent a powerful family of generative models widely used for image and video generation. However, the time-consuming deployment, long inference time, and requirements on large memory hinder their applications on resource constrained devices. In this paper, we propose a method based on the improved Straight-Through Estimator to improve the deployment efficiency of diffusion models. Specifically, we add sparse masks to the Convolution and Linear layers in a pre-trained diffusion model, then transfer learn the sparse model during the fine-tuning stage and turn on the sparse masks during inference. Experimental results on a Transformer and UNet-based diffusion models demonstrate that our method reduces MACs by 50% while maintaining FID. Sparse models are accelerated by approximately 1.2x on the GPU. Under other MACs conditions, the FID is also lower than 1 compared to other methods.

cs.LG↗

arXiv2025

Keeping Yourself is Important in Downstream Tuning Multimodal Large Language Model

Multi-modal Large Language Models (MLLMs) integrate visual and linguistic reasoning to address complex tasks such as image captioning and visual question answering. While MLLMs demonstrate remarkable versatility, MLLMs appears limited performance on special applications. But tuning MLLMs for downstream tasks encounters two key challenges: Task-Expert Specialization, where distribution shifts between pre-training and target datasets constrain target performance, and Open-World Stabilization, where catastrophic forgetting erases the model general knowledge. In this work, we systematically review recent advancements in MLLM tuning methodologies, classifying them into three paradigms: (I) Selective Tuning, (II) Additive Tuning, and (III) Reparameterization Tuning. Furthermore, we benchmark these tuning strategies across popular MLLM architectures and diverse downstream tasks to establish standardized evaluation analysis and systematic tuning principles. Finally, we highlight several open challenges in this domain and propose future research directions. To facilitate ongoing progress in this rapidly evolving field, we provide a public repository that continuously tracks developments: https://github.com/WenkeHuang/Awesome-MLLM-Tuning.

cs.CL↗

arXiv2025

Pulse Compression by an Optical Push Broom On a Chip

In this study, we report a first experimental demonstration of pulse compression by a gradual refractive index front moving in a periodically modulated silicon waveguide, the so-called optical push broom effect. Optical push broom captures and confines the input signal pulse in a faster propagating refractive index front, driven by a pump pulse. This is a spatio-temporal analogue of light trapping in a tapered plasmonic waveguide where light is continuously changing its wavevector approaching zero group velocity and, thus, stopped without reflection. Here the signal is accelerated by the front until the signal velocity matches the front velocity, thus stopping the light in respect to the front. We employ the slowly varying envelope approximation to model this phenomenon. Notably, we well reproduced the experimental frequency shift at the output corresponding to the temporal delay at the input.

physics.optics↗

arXiv2025

Cosmological Quasiparticles and the Cosmological Collider

The interplay between cosmology and strongly coupled dynamics can yield transient spectral features that vanish at late times, but which may leave behind phenomenological signatures in the spectrum of primordial fluctuations. Of particular interest are strongly coupled extensions of the standard model featuring approximate conformal invariance. In flat space, the spectral density for a scalar operator in a conformal field theory is characterized by a continuum with scaling law governed by the dimension of the operator, and is otherwise featureless. AdS/CFT arguments suggest that for large $N$, in an inflationary background with Hubble rate $H$, this continuum is gapped. We demonstrate that there can be additional peak structures that become sharp and particle-like at phenomenologically interesting regions in parameter space, and we estimate their contribution to cosmological observables. We find phenomena that are potentially observable in future experiments that are unique to these models, including displaced oscillatory features in the squeezed limit of the bi-spectrum. These particles can be either fundamental, and localized to a UV brane, or composite at the Hubble scale, $H$, and bound to a horizon in the bulk of the 5D geometry. We comment on how stabilization of conformal symmetry breaking vacua can be correlated with these spectral features and their phenomenology.

astro-ph.CO↗

arXiv2025

EmoAssist: Emotional Assistant for Visual Impairment Community

The rapid advancement of large multi-modality models (LMMs) has significantly propelled the integration of artificial intelligence into practical applications. Visual Question Answering (VQA) systems, which can process multi-modal data including vision, text, and audio, hold great potential for assisting the Visual Impairment (VI) community in navigating complex and dynamic real-world environments. However, existing VI assistive LMMs overlook the emotional needs of VI individuals, and current benchmarks lack emotional evaluation of these LMMs. To address these gaps, this paper introduces the EmoAssist Benchmark, a comprehensive benchmark designed to evaluate the assistive performance of LMMs for the VI community. To the best of our knowledge, this is the first benchmark that incorporates emotional intelligence as a key consideration. Furthermore, we propose the EmoAssist Model, an Emotion-Assistive LMM specifically designed for the VI community. The EmoAssist Model utilizes Direct Preference Optimization (DPO) to align outputs with human emotional preferences. Experiment results demonstrate that the EmoAssist Model significantly enhances the recognition of implicit emotions and intentions of VI users, delivers empathetic responses, and provides actionable guidance. Specifically, it shows respective improvements of 147.8% and 89.7% in the Empathy and Suggestion metrics on the EmoAssist Benchmark, compared to the pre-tuning LMM, and even outperforms state-of-the-art LLMs such as GPT-4o.

cs.CV↗

arXiv2024

Learn from Downstream and Be Yourself in Multimodal Large Language Model Fine-Tuning

Multimodal Large Language Model (MLLM) have demonstrated strong generalization capabilities across diverse distributions and tasks, largely due to extensive pre-training datasets. Fine-tuning MLLM has become a common practice to improve performance on specific downstream tasks. However, during fine-tuning, MLLM often faces the risk of forgetting knowledge acquired during pre-training, which can result in a decline in generalization abilities. To balance the trade-off between generalization and specialization, we propose measuring the parameter importance for both pre-trained and fine-tuning distributions, based on frozen pre-trained weight magnitude and accumulated fine-tuning gradient values. We further apply an importance-aware weight allocation strategy, selectively updating relatively important parameters for downstream tasks. We conduct empirical evaluations on both image captioning and visual question-answering tasks using various MLLM architectures. The comprehensive experimental analysis demonstrates the effectiveness of the proposed solution, highlighting the efficiency of the crucial modules in enhancing downstream specialization performance while mitigating generalization degradation in MLLM Fine-Tuning.

cs.CL↗

arXiv2024

Autoregressive Image Generation without Vector Quantization

Conventional wisdom holds that autoregressive models for image generation are typically accompanied by vector-quantized tokens. We observe that while a discrete-valued space can facilitate representing a categorical distribution, it is not a necessity for autoregressive modeling. In this work, we propose to model the per-token probability distribution using a diffusion procedure, which allows us to apply autoregressive models in a continuous-valued space. Rather than using categorical cross-entropy loss, we define a Diffusion Loss function to model the per-token probability. This approach eliminates the need for discrete-valued tokenizers. We evaluate its effectiveness across a wide range of cases, including standard autoregressive models and generalized masked autoregressive (MAR) variants. By removing vector quantization, our image generator achieves strong results while enjoying the speed advantage of sequence modeling. We hope this work will motivate the use of autoregressive generation in other continuous-valued domains and applications. Code is available at: https://github.com/LTH14/mar.

cs.CV↗

arXiv2024

Multi-Uncertainty Aware Autonomous Cooperative Planning

Autonomous cooperative planning (ACP) is a promising technique to improve the efficiency and safety of multi-vehicle interactions for future intelligent transportation systems. However, realizing robust ACP is a challenge due to the aggregation of perception, motion, and communication uncertainties. This paper proposes a novel multi-uncertainty aware ACP (MUACP) framework that simultaneously accounts for multiple types of uncertainties via regularized cooperative model predictive control (RC-MPC). The regularizers and constraints for perception, motion, and communication are constructed according to the confidence levels, weather conditions, and outage probabilities, respectively. The effectiveness of the proposed method is evaluated in the Car Learning to Act (CARLA) simulation platform. Results demonstrate that the proposed MUACP efficiently performs cooperative formation in real time and outperforms other benchmark approaches in various scenarios under imperfect knowledge of the environment.

cs.RO↗

arXiv2024

Continuous Approximations for Improving Quantization Aware Training of LLMs

Model compression methods are used to reduce the computation and energy requirements for Large Language Models (LLMs). Quantization Aware Training (QAT), an effective model compression method, is proposed to reduce performance degradation after quantization. To further minimize this degradation, we introduce two continuous approximations to the QAT process on the rounding function, traditionally approximated by the Straight-Through Estimator (STE), and the clamping function. By applying both methods, the perplexity (PPL) on the WikiText-v2 dataset of the quantized model reaches 9.0815, outperforming 9.9621 by the baseline. Also, we achieve a 2.76% improvement on BoolQ, and a 5.47% improvement on MMLU, proving that the step sizes and weights can be learned more accurately with our approach. Our method achieves better performance with the same precision, model size, and training setup, contributing to the development of more energy-efficient LLMs technology that aligns with global sustainability goals.

cs.LG↗

arXiv2024

Integer Topological Defects Reveal Anti-Symmetric Forces in Active Nematics

Cell layers are often categorized as contractile or extensile active nematics but recent experiments on neural progenitor cells with induced $+1$ topological defects challenge this classification. In a bottom-up approach, we first study a relevant particle-level model and then analyze a continuous theory derived from it. We show that both model and theory account qualitatively for the main experimental result, i.e. accumulation of cells at the core of any type of +1 defect. We argue that cell accumulation is essentially due to two generally ignored 'effective active forces'. We finally discuss the relevance and consequences of our findings in the context of other cellular active nematics experiments and previously proposed theories.

cond-mat.soft↗

arXiv2024

CipherDM: Secure Three-Party Inference for Diffusion Model Sampling

Diffusion Models (DMs) achieve state-of-the-art synthesis results in image generation and have been applied to various fields. However, DMs sometimes seriously violate user privacy during usage, making the protection of privacy an urgent issue. Using traditional privacy computing schemes like Secure Multi-Party Computation (MPC) directly in DMs faces significant computation and communication challenges. To address these issues, we propose CipherDM, the first novel, versatile and universal framework applying MPC technology to DMs for secure sampling, which can be widely implemented on multiple DM based tasks. We thoroughly analyze sampling latency breakdown, find time-consuming parts and design corresponding secure MPC protocols for computing nonlinear activations including SoftMax, SiLU and Mish. CipherDM is evaluated on popular architectures (DDPM, DDIM) using MNIST dataset and on SD deployed by diffusers. Compared to direct implementation on SPU, our approach improves running time by approximately 1.084\times \sim 2.328\times, and reduces communication costs by approximately 1.212\times \sim 1.791\times.

cs.CR↗

arXiv2024

Deep Band Crossings Enhanced Nonlinear Optical Effects

Nonlinear optical (NLO) effects in materials with band crossings have attracted significant research interests due to the divergent band geometric quantities around these crossings. Most current research has focused on band crossings between the valence and conduction bands. However, such crossings are absent in insulators, which are more relevant for NLO applications. In this work, we demonstrate that NLO effects can be significantly enhanced by band crossings within the valence or conduction bands, which we designate as "deep band crossings" (DBCs). As an example, in two dimensions, we show that shift conductivity can be substantially enhanced or even divergent due to a mirror-protected "deep Dirac nodal point". In three dimensions, we propose GeTe as an ideal material where shift conductivity is enhanced by "deep Dirac nodal lines". The ubiquity of this enhancement is further confirmed by high-throughput calculations. Other types of DBCs and NLO effects are also discussed. By manipulating band crossings between arbitrary bands, our work offers a simple, practical, and universal way to greatly enhance NLO effects.

cond-mat.mes-hall↗

arXiv2024

Neural-network Density Functional Theory Based on Variational Energy Minimization

Deep-learning density functional theory (DFT) shows great promise to significantly accelerate material discovery and potentially revolutionize materials research. However, current research in this field primarily relies on data-driven supervised learning, making the developments of neural networks and DFT isolated from each other. In this work, we present a theoretical framework of neural-network DFT, which unifies the optimization of neural networks with the variational computation of DFT, enabling physics-informed unsupervised learning. Moreover, we develop a differential DFT code incorporated with deep-learning DFT Hamiltonian, and introduce algorithms of automatic differentiation and backpropagation into DFT, demonstrating the capability of neural-network DFT. The physics-informed neural-network architecture not only surpasses conventional approaches in accuracy and efficiency, but also offers a new paradigm for developing deep-learning DFT methods.

physics.comp-ph↗

arXiv2024

Deep learning density functional theory Hamiltonian in real space

Deep learning electronic structures from ab initio calculations holds great potential to revolutionize computational materials studies. While existing methods proved success in deep-learning density functional theory (DFT) Hamiltonian matrices, they are limited to DFT programs using localized atomic-like bases and heavily depend on the form of the bases. Here, we propose the DeepH-r method for deep-learning DFT Hamiltonians in real space, facilitating the prediction of DFT Hamiltonian in a basis-independent manner. An equivariant neural network architecture for modeling the real-space DFT potential is developed, targeting a more fundamental quantity in DFT. The real-space potential exhibits simplified principles of equivariance and enhanced nearsightedness, further boosting the performance of deep learning. When applied to evaluate the Hamiltonian matrix, this method significantly improved in accuracy, as exemplified in multiple case studies. Given the abundance of data in the real-space potential, this work may pave a novel pathway for establishing a ``large materials model" with increased accuracy.

physics.comp-ph↗

arXiv2024

Direct ab initio calculation of the $^{4}$He nuclear electric dipole polarizability

The calculation of nuclear electromagnetic sum rules by directly diagonalizing the nuclear Hamiltonian in a large basis is numerically challenging and has not been performed for $A>2$ nuclei. With the significant progress of high performance computing, we show that calculating sum rules using numerous discretized continuum states obtained by directly diagonalizing the ab initio no-core shell model Hamiltonian is achievable numerically. Specifically, we calculate the $^{4}$He electric dipole ($E1$) polarizability, that is an inverse energy weighted sum rule, employing the Daejeon16 $NN$ interaction. We demonstrate that the calculations are numerically tractable as the dimension of the basis increases and are convergent. Our results for the $^{4}$He electric dipole polarizability are consistent with the most recent experimental data and are compared with those of other theoretical studies employing different techniques and various interactions.

nucl-th↗

arXiv2024

Look Within, Why LLMs Hallucinate: A Causal Perspective

The emergence of large language models (LLMs) is a milestone in generative artificial intelligence, achieving significant success in text comprehension and generation tasks. Despite the tremendous success of LLMs in many downstream tasks, they suffer from severe hallucination problems, posing significant challenges to the practical applications of LLMs. Most of the works about LLMs' hallucinations focus on data quality. Self-attention is a core module in transformer-based LLMs, while its potential relationship with LLMs' hallucination has been hardly investigated. To fill this gap, we study this problem from a causal perspective. We propose a method to intervene in LLMs' self-attention layers and maintain their structures and sizes intact. Specifically, we disable different self-attention layers in several popular open-source LLMs and then compare their degrees of hallucination with the original ones. We evaluate the intervened LLMs on hallucination assessment benchmarks and conclude that disabling some specific self-attention layers in the front or tail of the LLMs can alleviate hallucination issues. The study paves a new way for understanding and mitigating LLMs' hallucinations.

cs.CL↗