Active Preference Learning for Large Language Models

William Muldrew, Peter Hayes, Mingtian Zhang +1

As large language models (LLMs) become more capable, fine-tuning techniques for aligning with human intent are increasingly important. A key consideration for aligning these models…