Skip to yearly menu bar Skip to main content


Poster
in
Workshop: Accessible and Efficient Foundation Models for Biological Discovery

Graph2Token: Make LLMs Understand Molecule Graphs

Runze Wang · Mingqi Yang · Yanming Shen

Keywords: [ LLM token vocabulary ] [ Graph Tokenizer ] [ Lightweight Solution. ] [ Molecule Graph Token Alignment ]


Abstract:

Large language models (LLMs) excel at various text-related tasks. However, it is still challenging for them to process graph data such as molecules. To bridge this gap, this paper proposes Graph2Token, an efficient solution that aligns a graph token to LLM tokens. The key idea is to represent a graph token with the LLM token vocabulary, without finetuning the backbone of LLM. In this way, we can unleash the potential of existing LLMs, which helps the downstream molecule prediction tasks. Extensive experiments demonstrate the effectiveness of our proposed Graph2Token. Code is available athttps://anonymous.4open.science/r/Graph2Token.

Chat is not available.