SMILES

The 'SMILES' dataset contains molecular representations in the Simplified Molecular Input Line Entry System format and is used to evaluate the performance of large language models on various chemical tasks.

Papers using SMILES (7)

AReUReDi: Annealed Rectified Updates for Refining Discrete Flows with Multi-Objective Guidance2025 · 5 citesRethinking Molecular Text Representations for LLMs: An Empirical Study2026What an Autonomous Agent Discovers About Molecular Transformer Design: Does It Transfer?2026Synthesizable Molecular Generation via Soft-constrained GFlowNets with Rich Chemical Priors2026How to Make Large Language Models Generate 100% Valid Molecules?2025Can Hallucinations Help? Boosting LLMs for Drug Discovery2025Hierarchical Structure Enhances the Convergence and Generalizability of Linear Molecular Representation2024