Revisiting the Graph Reasoning Ability of Large Language Models: Case Studies in Translation, Connectivity and Shortest Path

TL;DR

This paper investigates the graph reasoning capabilities of Large Language Models across three fundamental tasks, revealing their limitations and variability in understanding graph structures through empirical evaluation and real-world data.

Contribution

It provides a comprehensive empirical analysis of LLMs' graph reasoning abilities, highlighting their failures and performance variability on key graph tasks.

Findings

LLMs struggle to understand graph structures from text descriptions.

Performance varies significantly across different graph reasoning tasks.

Real-world knowledge graphs show similar limitations as experimental results.

Abstract

Large Language Models (LLMs) have achieved great success in various reasoning tasks. In this work, we focus on the graph reasoning ability of LLMs. Although theoretical studies proved that LLMs are capable of handling graph reasoning tasks, empirical evaluations reveal numerous failures. To deepen our understanding on this discrepancy, we revisit the ability of LLMs on three fundamental graph tasks: graph description translation, graph connectivity, and the shortest-path problem. Our findings suggest that LLMs can fail to understand graph structures through text descriptions and exhibit varying performance for all these three fundamental tasks. Meanwhile, we perform a real-world investigation on knowledge graphs and make consistent observations with our findings. The codes and datasets are available.