MolDeTox

Summary

MolDeTox is a benchmark dataset designed to evaluate toxicity-aware molecular editing capabilities of LLMs and VLMs. The dataset is constructed based on the concept of *toxicity cliffs*, where structurally similar molecules exhibit opposite toxicity labels. This design enables models to learn minimal structural modifications for detoxification while preserving physicochemical properties.

Keywords

hf-dataset language-modeling tabular text molecule toxicity drug-discovery benchmark llm vlm

Topics

Language Modeling

Research notes

• downloads=139; likes=13