Large Language Models are In-Context Molecule Learners
Published in arXiv preprint, 2024
Recommended citation: Jiatong Li, Wei Liu, Zhihao Ding, Wenqi Fan, Yuqiang Li, Qing Li. (2024). Large Language Models are In-Context Molecule Learners. arXiv preprint arXiv:2403.04197. https://arxiv.org/pdf/2403.04197.pdf
Large Language Models (LLMs) have demonstrated exceptional performance in biochemical tasks, especially the molecule caption translation task, which aims to bridge the gap between molecules and natural language texts. However, previous methods in adapting LLMs to the molecule-caption translation task required extra domain-specific pre-training stages, suffered weak alignment between molecular and textual spaces, or imposed stringent demands on the scale of LLMs. To resolve the challenges, we propose In-Context Molecule Adaptation (ICMA) as a new paradigm allowing LLMs to learn the molecule-text alignment from con-text examples via In-Context Molecule Tuning. Specifically, ICMA incorporates the following three stages: Cross-modal Retrieval, Post-retrieval Re-ranking, and In-context Molecule Tuning. Initially, Cross-modal Retrieval utilises BM25 Caption Retrieval and Molecule Graph Retrieval to retrieve informative context examples. Additionally, we also propose Post-retrieval Re-ranking with Sequence Reversal and Random Walk to further improve the quality of retrieval results. Finally, In-Context Molecule Tuning unlocks the in-context molecule learning capability of LLMs with retrieved examples and adapts the parameters of LLMs for the molecule-caption translation task. Experimental results demonstrate that ICMT can empower LLMs to achieve state-of-the-art or comparable performance without extra training corpora and intricate structures, showing that LLMs are inherently in-context molecule learners.
Our codes will be available here: codes are available here
You could first download our model weighst via Huggingface Link