arXiv subjects

Yuhan Li

Publications and source records attributed to Yuhan Li.

At least 73 records · Page 4Linked to original sources

arXiv2025

GraphArena: Evaluating and Exploring Large Language Models on Graph Computation

The ``arms race'' of Large Language Models (LLMs) demands new benchmarks to examine their progresses. In this paper, we introduce GraphArena, a benchmarking tool designed to evaluate LLMs on real-world graph computational problems. It offers a suite of four polynomial-time tasks (e.g., Shortest Distance) and six NP-complete challenges (e.g., Traveling Salesman Problem). GraphArena features a rigorous evaluation framework that classifies LLM outputs as correct, suboptimal (feasible but not optimal), hallucinatory (properly formatted but infeasible), or missing. Evaluation of over 10 LLMs reveals that even top-performing LLMs struggle with larger, more complex graph problems and exhibit hallucination issues. We further explore four potential solutions to address this issue and improve LLMs on graph computation, including chain-of-thought prompting, instruction tuning, code writing, and scaling test-time compute, each demonstrating unique strengths and limitations. GraphArena complements the existing LLM benchmarks and is open-sourced at https://github.com/squareRoot3/GraphArena.

cs.AI↗

arXiv2024

GLBench: A Comprehensive Benchmark for Graph with Large Language Models

The emergence of large language models (LLMs) has revolutionized the way we interact with graphs, leading to a new paradigm called GraphLLM. Despite the rapid development of GraphLLM methods in recent years, the progress and understanding of this field remain unclear due to the lack of a benchmark with consistent experimental protocols. To bridge this gap, we introduce GLBench, the first comprehensive benchmark for evaluating GraphLLM methods in both supervised and zero-shot scenarios. GLBench provides a fair and thorough evaluation of different categories of GraphLLM methods, along with traditional baselines such as graph neural networks. Through extensive experiments on a collection of real-world datasets with consistent data processing and splitting strategies, we have uncovered several key findings. Firstly, GraphLLM methods outperform traditional baselines in supervised settings, with LLM-as-enhancers showing the most robust performance. However, using LLMs as predictors is less effective and often leads to uncontrollable output issues. We also notice that no clear scaling laws exist for current GraphLLM methods. In addition, both structures and semantics are crucial for effective zero-shot transfer, and our proposed simple baseline can even outperform several models tailored for zero-shot scenarios. The data and code of the benchmark can be found at https://github.com/NineAbyss/GLBench.

cs.LG↗

arXiv2024

GCoder: Improving Large Language Model for Generalized Graph Problem Solving

Large Language Models (LLMs) have demonstrated strong reasoning abilities, making them suitable for complex tasks such as graph computation. Traditional reasoning steps paradigm for graph problems is hindered by unverifiable steps, limited long-term reasoning, and poor generalization to graph variations. To overcome these limitations, we introduce GCoder, a code-based LLM designed to enhance problem-solving in generalized graph computation problems. Our method involves constructing an extensive training dataset, GraphWild, featuring diverse graph formats and algorithms. We employ a multi-stage training process, including Supervised Fine-Tuning (SFT) and Reinforcement Learning from Compiler Feedback (RLCF), to refine model capabilities. For unseen tasks, a hybrid retrieval technique is used to augment performance. Experiments demonstrate that GCoder outperforms GPT-4o, with an average accuracy improvement of 16.42% across various graph computational problems. Furthermore, GCoder efficiently manages large-scale graphs with millions of nodes and diverse input formats, overcoming the limitations of previous models focused on the reasoning steps paradigm. This advancement paves the way for more intuitive and effective graph problem-solving using LLMs. Code and data are available at here: https://github.com/Bklight999/WWW25-GCoder/tree/master.

cs.CL↗

arXiv2024

GraphWiz: An Instruction-Following Language Model for Graph Problems

Large language models (LLMs) have achieved impressive success across several fields, but their proficiency in understanding and resolving complex graph problems is less explored. To bridge this gap, we introduce GraphInstruct, a novel and comprehensive instruction-tuning dataset designed to equip language models with the ability to tackle a broad spectrum of graph problems using explicit reasoning paths. Utilizing GraphInstruct, we build GraphWiz, an open-source language model capable of resolving various graph problem types while generating clear reasoning processes. To enhance the model's capability and reliability, we incorporate the Direct Preference Optimization (DPO) framework into the graph problem-solving context. The enhanced model, GraphWiz-DPO, achieves an average accuracy of 65% across nine tasks with different complexity levels, surpassing GPT-4 which has an average accuracy of 43.8%. Moreover, our research delves into the delicate balance between training data volume and model performance, highlighting the potential for overfitting with increased data. We also explore the transferability of the model's reasoning ability across different graph tasks, indicating the model's adaptability and practical application potential. Our investigation offers a new blueprint and valuable insights for developing LLMs specialized in graph reasoning and problem-solving.

cs.CL↗

arXiv2024

ZeroG: Investigating Cross-dataset Zero-shot Transferability in Graphs

With the development of foundation models such as large language models, zero-shot transfer learning has become increasingly significant. This is highlighted by the generative capabilities of NLP models like GPT-4, and the retrieval-based approaches of CV models like CLIP, both of which effectively bridge the gap between seen and unseen data. In the realm of graph learning, the continuous emergence of new graphs and the challenges of human labeling also amplify the necessity for zero-shot transfer learning, driving the exploration of approaches that can generalize across diverse graph data without necessitating dataset-specific and label-specific fine-tuning. In this study, we extend such paradigms to zero-shot transferability in graphs by introducing ZeroG, a new framework tailored to enable cross-dataset generalization. Addressing the inherent challenges such as feature misalignment, mismatched label spaces, and negative transfer, we leverage a language model to encode both node attributes and class semantics, ensuring consistent feature dimensions across datasets. We also propose a prompt-based subgraph sampling module that enriches the semantic information and structure information of extracted subgraphs using prompting nodes and neighborhood aggregation, respectively. We further adopt a lightweight fine-tuning strategy that reduces the risk of overfitting and maintains the zero-shot learning efficacy of the language model. The results underscore the effectiveness of our model in achieving significant cross-dataset zero-shot transferability, opening pathways for the development of graph foundation models. Codes and data are available at https://github.com/NineAbyss/ZeroG.

cs.LG↗

arXiv2024

EDGE-LLM: Enabling Efficient Large Language Model Adaptation on Edge Devices via Layerwise Unified Compression and Adaptive Layer Tuning and Voting

Efficient adaption of large language models (LLMs) on edge devices is essential for applications requiring continuous and privacy-preserving adaptation and inference. However, existing tuning techniques fall short because of the high computation and memory overheads. To this end, we introduce a computation- and memory-efficient LLM tuning framework, called Edge-LLM, to facilitate affordable and effective LLM adaptation on edge devices. Specifically, Edge-LLM features three core components: (1) a layer-wise unified compression (LUC) technique to reduce the computation overhead by generating layer-wise pruning sparsity and quantization bit-width policies, (2) an adaptive layer tuning and voting scheme to reduce the memory overhead by reducing the backpropagation depth, and (3) a complementary hardware scheduling strategy to handle the irregular computation patterns introduced by LUC and adaptive layer tuning, thereby achieving efficient computation and data movements. Extensive experiments demonstrate that Edge-LLM achieves a 2.92x speed up and a 4x memory overhead reduction as compared to vanilla tuning methods with comparable task accuracy. Our code is available at https://github.com/GATECH-EIC/Edge-LLM

cs.LG↗

arXiv2024

The impact of nodes of information dissemination on epidemic spreading in dynamic multiplex networks

Epidemic spreading processes on dynamic multiplex networks provide a more accurate description of natural spreading processes than those on single layered networks. To describe the influence of different individuals in the awareness layer on epidemic spreading, we propose a two-layer network-based epidemic spreading model, including some individuals who neglect the epidemic, and we explore how individuals with different properties in the awareness layer will affect the spread of epidemics. The two-layer network model is divided into an information transmission layer and a disease spreading layer. Each node in the layer represents an individual with different connections in different layers. Individuals with awareness will be infected with a lower probability compared to unaware individuals, which corresponds to the various epidemic prevention measures in real life. We adopt the micro-Markov chain approach to analytically derive the threshold for the proposed epidemic model, which demonstrates that the awareness layer affects the threshold of disease spreading. We then explore how individuals with different properties would affect the disease spreading process through extensive Monte Carlo numerical simulations. We find that individuals with high centrality in the awareness layer would significantly inhibit the transmission of infectious diseases. Additionally, we propose conjectures and explanations for the approximately linear effect of individuals with low centrality in the awareness layer on the number of infected individuals.

cs.SI↗

arXiv2024

A Seamless Phase II/III Design with Dose Optimization for Oncology Drug Development

The US FDA's Project Optimus initiative that emphasizes dose optimization prior to marketing approval represents a pivotal shift in oncology drug development. It has a ripple effect for rethinking what changes may be made to conventional pivotal trial designs to incorporate a dose optimization component. Aligned with this initiative, we propose a novel Seamless Phase II/III Design with Dose Optimization (SDDO framework). The proposed design starts with dose optimization in a randomized setting, leading to an interim analysis focused on optimal dose selection, trial continuation decisions, and sample size re-estimation (SSR). Based on the decision at interim analysis, patient enrollment continues for both the selected dose arm and control arm, and the significance of treatment effects will be determined at final analysis. The SDDO framework offers increased flexibility and cost-efficiency through sample size adjustment, while stringently controlling the Type I error. This proposed design also facilitates both Accelerated Approval (AA) and regular approval in a "one-trial" approach. Extensive simulation studies confirm that our design reliably identifies the optimal dosage and makes preferable decisions with a reduced sample size while retaining statistical power.

stat.ME↗

arXiv2024

Effect of antibody levels on the spread of disease in multiple infections

There are complex interactions between antibody levels and epidemic propagation, the antibody level of an individual influences the probability of infection, and the spread of the virus influences the antibody level of each individual. There exist some viruses that, in their natural state, cause antibody levels in an infected individual to gradually decay. When these antibody levels decay to a certain point, the individual can be reinfected, such as with COVID 19. To describe their interaction, we introduce a novel mathematical model that incorporates the presence of an antibody retention rate to investigate the infection patterns of individuals who survive multiple infections. The model is composed of a system of stochastic differential equations to derive the equilibrium point and threshold of the model and presents rich experimental results of numerical simulations to further elucidate the propagation properties of the model. We find that the antibody decay rate strongly affects the propagation process, and also that different network structures have different sensitivities to the antibody decay rate, and that changes in the antibody decay rate cause stronger changes in the propagation process in Barabasi Albert networks. Furthermore, we investigate the stationary distribution of the number of infection states and the final antibody levels, and find that they both satisfy the normal distribution, but the standard deviation is small in the Barabasi Albert network. Finally, we explore the effect of individual antibody differences and decay rates on the final population antibody levels, and uncover that individual antibody differences do not affect the final mean antibody levels. The study offers valuable insights for epidemic prevention and control in practical applications.

q-bio.PE↗

arXiv2024

AnyFit: Controllable Virtual Try-on for Any Combination of Attire Across Any Scenario

While image-based virtual try-on has made significant strides, emerging approaches still fall short of delivering high-fidelity and robust fitting images across various scenarios, as their models suffer from issues of ill-fitted garment styles and quality degrading during the training process, not to mention the lack of support for various combinations of attire. Therefore, we first propose a lightweight, scalable, operator known as Hydra Block for attire combinations. This is achieved through a parallel attention mechanism that facilitates the feature injection of multiple garments from conditionally encoded branches into the main network. Secondly, to significantly enhance the model's robustness and expressiveness in real-world scenarios, we evolve its potential across diverse settings by synthesizing the residuals of multiple models, as well as implementing a mask region boost strategy to overcome the instability caused by information leakage in existing models. Equipped with the above design, AnyFit surpasses all baselines on high-resolution benchmarks and real-world data by a large gap, excelling in producing well-fitting garments replete with photorealistic and rich details. Furthermore, AnyFit's impressive performance on high-fidelity virtual try-ons in any scenario from any image, paves a new path for future research within the fashion community.

cs.CV↗

arXiv2024

Approximating Multi-Dimensional and Multiband Signals

We study the problem of representing a discrete tensor that comes from finite uniform samplings of a multi-dimensional and multiband analog signal. Particularly, we consider two typical cases in which the shape of the subbands is cubic or parallelepipedic. For the cubic case, by examining the spectrum of its corresponding time- and band-limited operators, we obtain a low-dimensional optimal dictionary to represent the original tensor. We further prove that the optimal dictionary can be approximated by the famous \ac{dpss} with certain modulation, leading to an efficient constructing method. For the parallelepipedic case, we show that there also exists a low-dimensional dictionary to represent the original tensor. We present rigorous proof that the numbers of atoms in both dictionaries are approximately equal to the dot of the total number of samplings and the total volume of the subbands. Our derivations are mainly focused on the \ac{2d} scenarios but can be naturally extended to high dimensions.

eess.SP↗

arXiv2024

Image to Pseudo-Episode: Boosting Few-Shot Segmentation by Unlabeled Data

Few-shot segmentation (FSS) aims to train a model which can segment the object from novel classes with a few labeled samples. The insufficient generalization ability of models leads to unsatisfactory performance when the models lack enough labeled data from the novel classes. Considering that there are abundant unlabeled data available, it is promising to improve the generalization ability by exploiting these various data. For leveraging unlabeled data, we propose a novel method, named Image to Pseudo-Episode (IPE), to generate pseudo-episodes from unlabeled data. Specifically, our method contains two modules, i.e., the pseudo-label generation module and the episode generation module. The former module generates pseudo-labels from unlabeled images by the spectral clustering algorithm, and the latter module generates pseudo-episodes from pseudo-labeled images by data augmentation methods. Extensive experiments on PASCAL-$5^i$ and COCO-$20^i$ demonstrate that our method achieves the state-of-the-art performance for FSS.

cs.CV↗

arXiv2024

A Survey of Graph Meets Large Language Model: Progress and Future Directions

Graph plays a significant role in representing and analyzing complex relationships in real-world applications such as citation networks, social networks, and biological data. Recently, Large Language Models (LLMs), which have achieved tremendous success in various domains, have also been leveraged in graph-related tasks to surpass traditional Graph Neural Networks (GNNs) based methods and yield state-of-the-art performance. In this survey, we first present a comprehensive review and analysis of existing methods that integrate LLMs with graphs. First of all, we propose a new taxonomy, which organizes existing methods into three categories based on the role (i.e., enhancer, predictor, and alignment component) played by LLMs in graph-related tasks. Then we systematically survey the representative methods along the three categories of the taxonomy. Finally, we discuss the remaining limitations of existing studies and highlight promising avenues for future research. The relevant papers are summarized and will be consistently updated at: https://github.com/yhLeeee/Awesome-LLMs-in-Graph-tasks.

cs.LG↗

arXiv2024

On inverse problems in multi-population aggregation models

This paper focuses on inverse problems arising in studying multi-population aggregations. The goal is to reconstruct the diffusion coefficient, advection coefficient, and interaction kernels of the aggregation system, which characterize the dynamics of different populations. In the theoretical analysis of the physical setup, it is crucial to ensure non-negativity of solutions. To address this, we employ the high-order variation method and introduce modifications to the systems. Additionally, we propose a novel approach called transformative asymptotic technique that enables the recovery of the diffusion coefficient preceding the Laplace operator, presenting a pioneering method for this type of problems. Through these techniques, we offer comprehensive insights into the unique identifiability aspect of inverse problems associated with multi-population aggregation models.

math.AP↗

arXiv2023

All Data on the Table: Novel Dataset and Benchmark for Cross-Modality Scientific Information Extraction

Extracting key information from scientific papers has the potential to help researchers work more efficiently and accelerate the pace of scientific progress. Over the last few years, research on Scientific Information Extraction (SciIE) witnessed the release of several new systems and benchmarks. However, existing paper-focused datasets mostly focus only on specific parts of a manuscript (e.g., abstracts) and are single-modality (i.e., text- or table-only), due to complex processing and expensive annotations. Moreover, core information can be present in either text or tables or across both. To close this gap in data availability and enable cross-modality IE, while alleviating labeling costs, we propose a semi-supervised pipeline for annotating entities in text, as well as entities and relations in tables, in an iterative procedure. Based on this pipeline, we release novel resources for the scientific community, including a high-quality benchmark, a large-scale corpus, and a semi-supervised annotation pipeline. We further report the performance of state-of-the-art IE models on the proposed benchmark dataset, as a baseline. Lastly, we explore the potential capability of large language models such as ChatGPT for the current task. Our new dataset, results, and analysis validate the effectiveness and efficiency of our semi-supervised pipeline, and we discuss its remaining limitations.

cs.CL↗

arXiv2023

On inverse problems in predator-prey models

In this paper, we consider the inverse problem of determining the coefficients of interaction terms within some Lotka-Volterra models, with support from boundary observation of its non-negative solutions. In the physical background, the solutions to the predator-prey model stand for the population densities for predator and prey and are non-negative, which is a critical challenge in our inverse problem study. We mainly focus on the unique identifiability issue and tackle it with the high-order variation method, a relatively new technique introduced by the second author and his collaborators. This method can ensure the positivity of solutions and has broader applicability in other physical models with non-negativity requirements. Our study improves this method by choosing a more general solution $(u_0,v_0)$ to expand around, achieving recovery for all interaction terms. By this means, we improve on the previous results and apply this to physical models to recover coefficients concerning compression, prey attack, crowding, carrying capacity, and many other interaction factors in the system. Finally, we apply our results to study three specific cases: the hydra-effects model, the Holling-Tanner model and the classic Lotka-Volterra model.

math.AP↗

arXiv2023

AI in Pharma for Personalized Sequential Decision-Making: Methods, Applications and Opportunities

In the pharmaceutical industry, the use of artificial intelligence (AI) has seen consistent growth over the past decade. This rise is attributed to major advancements in statistical machine learning methodologies, computational capabilities and the increased availability of large datasets. AI techniques are applied throughout different stages of drug development, ranging from drug discovery to post-marketing benefit-risk assessment. Kolluri et al. provided a review of several case studies that span these stages, featuring key applications such as protein structure prediction, success probability estimation, subgroup identification, and AI-assisted clinical trial monitoring. From a regulatory standpoint, there was a notable uptick in submissions incorporating AI components in 2021. The most prevalent therapeutic areas leveraging AI were oncology (27%), psychiatry (15%), gastroenterology (12%), and neurology (11%). The paradigm of personalized or precision medicine has gained significant traction in recent research, partly due to advancements in AI techniques \cite{hamburg2010path}. This shift has had a transformative impact on the pharmaceutical industry. Departing from the traditional "one-size-fits-all" model, personalized medicine incorporates various individual factors, such as environmental conditions, lifestyle choices, and health histories, to formulate customized treatment plans. By utilizing sophisticated machine learning algorithms, clinicians and researchers are better equipped to make informed decisions in areas such as disease prevention, diagnosis, and treatment selection, thereby optimizing health outcomes for each individual.

stat.ME↗

arXiv2023

Large Language Models Meet Harry Potter: A Bilingual Dataset for Aligning Dialogue Agents with Characters

In recent years, Dialogue-style Large Language Models (LLMs) such as ChatGPT and GPT4 have demonstrated immense potential in constructing open-domain dialogue agents. However, aligning these agents with specific characters or individuals remains a considerable challenge due to the complexities of character representation and the lack of comprehensive annotations. In this paper, we introduce the Harry Potter Dialogue (HPD) dataset, designed to advance the study of dialogue agents and character alignment. The dataset encompasses all dialogue sessions (in both English and Chinese) from the Harry Potter series and is annotated with vital background information, including dialogue scenes, speakers, character relationships, and attributes. These extensive annotations may empower LLMs to unlock character-driven dialogue capabilities. Furthermore, it can serve as a universal benchmark for evaluating how well can a LLM aligning with a specific character. We benchmark LLMs on HPD using both fine-tuning and in-context learning settings. Evaluation results reveal that although there is substantial room for improvement in generating high-quality, character-aligned responses, the proposed dataset is valuable in guiding models toward responses that better align with the character of Harry Potter.

cs.CL↗