Extended Data Table 3 Performances of different LLMs on three new datasets

From: A collaborative large language model for drug analysis

  1. The three new datasets, that is, DrugBank-QA, MIMIC-DrugQA, and COVID-Moderna, can be ensured to have not been used during the training process of current LLMs. The experiment validates the performance of the LLMs without potential data leakage and examines whether the LLMs can provide decision support for new drugs. We report the average performance. Bold values indicate the highest performance for each dataset.
Back to article page