Awesome RLHF is a curated collection of research papers and resources focused on reinforcement learning with human feedback.
The collection addresses the need to track and organize the growing body of work in RLHF, a technique that uses human feedback to align language models and other agents with human values and preferences. The repository organizes papers chronologically and provides an overview of how RLHF works in different domains, from large language models to video games, alongside explanations of related techniques like inverse reinforcement learning and apprenticeship learning.
Developers and researchers studying RLHF should use this resource to survey the landscape of published work in the field. The repository structures papers by year and includes sections for codebases, datasets, blogs, and books, making it useful for those implementing RLHF systems or conducting literature reviews. It serves as a reference point for understanding both foundational concepts and recent advances in the area.
The project is actively maintained with continuous updates to track new research as it emerges in the RLHF space. The repository accepts community contributions to expand its coverage of papers and resources.
