This dataset is designed for RLVR (Reinforcement Learning with Verifiable Rewards) tasks focused on molecular geometry manipulation. It enables training AI agents to programmatically manipulate molecular geometries like a computational chemist.
Dataset Description
The dataset contains molecular conformations in XYZ format with corresponding SMILES strings, sourced from the GeomConf dataset.
Dataset Statistics… See the full description on the dataset page: https://huggingface.co/datasets/yunhengzounvidia/geomconf-rlvr.
