Model-based Preference Optimization in Abstractive Summarization without Human Feedback

Open in new window