Improving language model behavior by training on a curated dataset

By /
Our latest research finds we can improve language model behavior with respect to specific behavioral values by fine-tuning on a small, curated dataset.
Scroll to Top