arxiv:2511.11257

AIonopedia: an LLM agent orchestrating multimodal learning for ionic liquid discovery

Published on Nov 14

· Submitted by

Yuqi Yin on Nov 17

Upvote

Authors:

Yuqi Yin ,

Siyuan Wang ,

Peng Sun ,

Abstract

Axonopedia, an AI agent utilizing LLMs and a multimodal foundation model, enhances property prediction and molecular design for Ionic Liquids through hierarchical search and real-world validation.

AI-generated summary

The discovery of novel Ionic Liquids (ILs) is hindered by critical challenges in property prediction, including limited data, poor model accuracy, and fragmented workflows. Leveraging the power of Large Language Models (LLMs), we introduce AIonopedia, to the best of our knowledge, the first LLM agent for IL discovery. Powered by an LLM-augmented multimodal domain foundation model for ILs, AIonopedia enables accurate property predictions and incorporates a hierarchical search architecture for molecular screening and design. Trained and evaluated on a newly curated and comprehensive IL dataset, our model delivers superior performance. Complementing these results, evaluations on literature-reported systems indicate that the agent can perform effective IL modification. Moving beyond offline tests, the practical efficacy was further confirmed through real-world wet-lab validation, in which the agent demonstrated exceptional generalization capabilities on challenging out-of-distribution tasks, underscoring its ability to accelerate real-world IL discovery.

View arXiv page View PDF Add to collection

Community

chitanda-eru

Paper author Paper submitter 3 days ago

We introduce AIonopedia, to our knowledge the first efficient LLM-based intelligent agent tailored to ionic liquids (ILs). By interacting with various specialized modules, it orchestrates the execution of multiple IL-related pipelines. At its core is an LLM-augmented multimodal domain foundation model for IL property prediction. To support this model, we compile a new IL dataset containing the largest collection to date of known IL solute-solvent interaction data. The resulting system achieves consistently strong performance across diverse IL property benchmarks, including robust out-of-distribution generalization. Building on these capabilities, AIonopedia offers automated pipelines for IL modification and large-scale molecular screening, which we further validate through wet-lab experiments, demonstrating its practical utility for data-driven IL research and design.