All publications
Journal of Chemical Information and Modeling2026Journal

RxnBench: A Multimodal Benchmark for Evaluating Large Language Models on Chemical Reaction Understanding from the Scientific Literature

Hanzheng Li, Xi Fang, Yixuan Li, Chaozheng Huang, Junjie Wang, Xi Wang, Hongzhe Bai, Bojun Hao, Shenyu Lin, Huiqi Liang, Linfeng Zhang, Guolin Ke

Read paper

Abstract

RxnBench evaluates multimodal language models on chemical-reaction understanding grounded in scientific literature. Its single-figure and full-document tasks test visual perception, mechanistic reasoning, and cross-modal synthesis across text, reaction schemes, and tables.

In brief

A multimodal benchmark for evaluating how well language models understand chemical reactions in scientific literature.