Chem4DLLM: 4D Multimodal LLMs for Chemical Dynamics Understanding
This addresses the problem for chemists needing to model dynamic chemical phenomena like bond breaking, though it is incremental as it builds on existing multimodal LLM approaches.
The authors tackled the limitation of static molecular representations in chemical understanding by introducing Chemical Dynamics Understanding (ChemDU), a task that translates 4D molecular trajectories into natural-language explanations, and developed Chem4DLLM, a model integrating an equivariant graph encoder with a pretrained LLM to achieve this.
Existing chemical understanding tasks primarily rely on static molecular representations, limiting their ability to model inherently dynamic phenomena such as bond breaking or conformational changes, which are essential for a chemist to understand chemical reactions. To address this gap, we introduce Chemical Dynamics Understanding (ChemDU), a new task that translates 4D molecular trajectories into interpretable natural-language explanations. ChemDU focuses on fundamental dynamic scenarios, including gas-phase and catalytic reactions, and requires models to reason about key events along molecular trajectories, such as bond formation and dissociation, and to generate coherent, mechanistically grounded narratives. To benchmark this capability, we construct Chem4DBench, the first dataset pairing 4D molecular trajectories with expert-authored explanations across these settings. We further propose Chem4DLLM, a unified model that integrates an equivariant graph encoder with a pretrained large language model to explicitly capture molecular geometry and rotational dynamics. We hope that ChemDU, together with Chem4DBench and Chem4DLLM, will stimulate further research in dynamic chemical understanding and multimodal scientific reasoning.