Saved in:
Bibliographic Details
Main Authors: Wang, Yuxiang, Dai, Xinnan, Fan, Wenqi, Ma, Yao
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2502.18771
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915756940197888
author Wang, Yuxiang
Dai, Xinnan
Fan, Wenqi
Ma, Yao
author_facet Wang, Yuxiang
Dai, Xinnan
Fan, Wenqi
Ma, Yao
contents In recent years, large language models (LLMs) have emerged as promising candidates for graph tasks. Many studies leverage natural language to describe graphs and apply LLMs for reasoning, yet most focus narrowly on performance benchmarks without fully comparing LLMs to graph learning models or exploring their broader potential. In this work, we present a comprehensive study of LLMs on graph learning tasks, evaluating both off-the-shelf and instruction-tuned models across a variety of scenarios. Beyond accuracy, we discuss data leakage concerns and computational overhead, and assess their performance under few-shot/zero-shot settings, domain transfer, structural understanding, and robustness. Our findings show that LLMs, particularly those with instruction tuning, greatly outperform traditional graph learning models in few-shot settings, exhibit strong domain transferability, and demonstrate excellent generalization and robustness. Our study highlights the broader capabilities of LLMs in graph learning and provides a foundation for future research.
format Preprint
id arxiv_https___arxiv_org_abs_2502_18771
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Exploring Graph Learning Tasks with Pure LLMs: A Comprehensive Benchmark and Investigation
Wang, Yuxiang
Dai, Xinnan
Fan, Wenqi
Ma, Yao
Machine Learning
Social and Information Networks
In recent years, large language models (LLMs) have emerged as promising candidates for graph tasks. Many studies leverage natural language to describe graphs and apply LLMs for reasoning, yet most focus narrowly on performance benchmarks without fully comparing LLMs to graph learning models or exploring their broader potential. In this work, we present a comprehensive study of LLMs on graph learning tasks, evaluating both off-the-shelf and instruction-tuned models across a variety of scenarios. Beyond accuracy, we discuss data leakage concerns and computational overhead, and assess their performance under few-shot/zero-shot settings, domain transfer, structural understanding, and robustness. Our findings show that LLMs, particularly those with instruction tuning, greatly outperform traditional graph learning models in few-shot settings, exhibit strong domain transferability, and demonstrate excellent generalization and robustness. Our study highlights the broader capabilities of LLMs in graph learning and provides a foundation for future research.
title Exploring Graph Learning Tasks with Pure LLMs: A Comprehensive Benchmark and Investigation
topic Machine Learning
Social and Information Networks
url https://arxiv.org/abs/2502.18771