R. R. Meier

About

R. R. Meier is a rising researcher at the forefront of unsupervised reinforcement learning, specializing in the development of intrinsically motivated agents that learn without external rewards. Their most notable contribution, the 2022 paper “Open-Ended Reinforcement Learning with Neural Reward Functions,” tackles a fundamental challenge in AI: enabling agents to autonomously discover diverse, reusable skills in complex environments. By proposing a novel framework that replaces handcrafted reward functions with learned, neural-based intrinsic rewards, Meier’s work extends the principles of unsupervised learning—so successful in computer vision and NLP—into the reinforcement learning domain. This approach moves beyond popular methods like DIAYN and DADS, which optimize mutual information, by fostering more open-ended and scalable skill discovery. Although early in their career, with the paper already garnering 3 citations, Meier’s research is gaining traction for its potential to unlock more general and adaptable AI systems. Their work is particularly relevant for students and researchers interested in the intersection of reinforcement learning, intrinsic motivation, and lifelong learning, offering a fresh perspective on how agents can learn continuously without predefined goals.

Research Focus

Key Achievements

Top Papers

1. 1
   Open-Ended Reinforcement Learning with Neural Reward Functions

   3 citations · 2022

Key Collaborators

Contact & Links