The Curious Case of Factuality Finetuning: Models' Internal Beliefs Can Improve Factuality

Open in new window