Reinforcement learning from human feedback

In machine learning, reinforcement learning from human feedback (RLHF) or reinforcement learning from human preferences is a technique that trains a "reward model" directly from human feedback and uses the model as a reward function to optimize an agent's policy using reinforcement learning (RL) through an optimization algorithm like Proximal Policy Optimization.[1][2] The reward model is trained in advance to the policy being optimized to predict if a given output is good (high reward) or bad (low reward). RLHF can improve the robustness and exploration of RL agents, especially when the reward function is sparse or noisy.[3]

Human feedback is collected by asking humans to rank instances of the agent's behavior.[4][5][6] These rankings can then be used to score outputs, for example with the Elo rating system.[2]

RLHF has been applied to various domains of natural language processing, such as conversational agents, text summarization, and natural language understanding.[7] Ordinary reinforcement learning, where agents learn from their own actions based on a "reward function", is difficult to apply to natural language processing tasks because the rewards are often not easy to define or measure, especially when dealing with complex tasks that involve human values or preferences. RLHF can enable language models to provide answers that align with these complex values, to generate more verbose responses, and to reject questions that are either inappropriate or outside the knowledge space of the model.[8] Some examples of RLHF-trained language models are OpenAI's ChatGPT and its predecessor InstructGPT,[5][9] as well as DeepMind's Sparrow.[10]

RLHF has also been applied to other areas, such as the development of video game bots. For example, OpenAI and DeepMind trained agents to play Atari games based on human preferences.[11][12] The agents achieved strong performance in many of the environments tested, often surpassing human performance.[13]