Principle-Driven Self-Alignment of Language Models from Scratch with Minimal Human Supervision

By Zhiqing Sun, Yikang Shen, Qinhong Zhou, Hongxin Zhang, Zhenfang Chen, David D. Cox, Yiming Yang, Chuang Gan. 2023

457 citations recorded by the source corpus.

Recent AI-assistant agents, such as ChatGPT, predominantly rely on supervised fine-tuning (SFT) with human annotations and reinforcement learning from human feedback (RLHF) to align the output of large language models (LLMs) with human intentions, ensuring they are helpful, ethical, and reliable. However, this dependence can significantly constrain the true potential of AI-assistant agents due to the high cost of obtaining human supervision and the related issues on quality, reliability, diversity, self-consistency, and undesirable biases. To address these challenges, we propose a novel approach called SELF-ALIGN, which combines principle-driven reasoning and the generative power of LLMs for the self-alignment of AI agents with minimal human supervision. Our approach encompasses four stages: first, we use an LLM to generate synthetic prompts, and a topic-guided method to augment the prompt diversity; second, we use a small set of human-written principles for AI models to follow, and guide the LLM through in-context learning from demonstrations (of principles application) to produce helpful, ethical, and reliable responses to user's queries; third, we fine-tune the original LLM with the high-quality self-aligned responses so that the resulting model can generate desirable responses for each query directly without the principle set and the demonstrations anymore; and finally, we offer a refinement step to address the issues of overly-brief or indirect responses. Applying SELF-ALIGN to the LLaMA-65b base language model, we develop an AI assistant named Dromedary. With fewer than 300 lines of human annotations (including<200 seed prompts, 16 generic principles, and 5 exemplars for in-context learning). Dromedary significantly surpasses the performance of several state-of-the-art AI systems, including Text-Davinci-003 and Alpaca, on benchmark datasets with various settings.

Subfields: Artificial Intelligence, Large Language Models, Alignment & RLHF

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

  1. Mohit Bansal
  2. Luke S. Zettlemoyer
  3. Graham Neubig
  4. Yejin Choi
  5. Percy Liang
  6. Hannaneh Hajishirzi
  7. Noah A. Smith
  8. Jianfeng Gao

In Large Language Models

  1. Yejin Choi
  2. Luke S. Zettlemoyer
  3. Noah A. Smith
  4. Hannaneh Hajishirzi
  5. Michael Lewis
  6. Zhiyuan Liu
  7. Bill Yuchen Lin
  8. Maosong Sun

In Alignment & RLHF

  1. Luke S. Zettlemoyer
  2. Yejin Choi
  3. Alexander Wei
  4. Noah A. Smith
  5. Yu Meng
  6. Yilun Du
  7. Swaroop Mishra
  8. Hannaneh Hajishirzi

Citations in the corpus