École Nationale Supérieure de
Techniques Avancées

Abstract

This work explores the application of large language models (LLMs) in graph generation and addresses the graph isomorphism problem. By developing a novel graph generative model (GGM) and utilizing advanced prompt engineering techniques, the study compares the effectiveness of LLMs with the new model. Additionally, it introduces a unique approach to solving the graph isomorphism problem using breadth-first search (BFS) and dynamic programming (DP). The research contributes to advancements in graph theory, offering insights into both automated graph processing and the theoretical aspects of graph isomorphism.