Anthropic Fellows Program, Reinforcement Learning
Core
Empirical AI research project focused on Reinforcement Learning, utilizing external infrastructure to produce public outputs like papers.
Role type
Research Fellow (Reinforcement Learning)
Builds
RL environments, model-based tools for training data analysis, and solutions for RL algorithms.
Domain
Artificial Intelligence / Machine Learning / Reinforcement Learning
Deliverable
research
Required skills
Python programming, software engineering for complex ML systems, training/fine-tuning/evaluating large language models, analyzing and debugging model training processes, working with large-scale distributed systems and high-performance computing
Preferred skills
Experience in AI safety/security, economics, or social sciences; experience in research or engineering related to the specific workstream
Technologies
Python, large-scale distributed systems, high-performance computing
Responsibilities
Conduct empirical research on RL algorithms and environments; implement solutions to improve model capabilities or safety; collaborate across research and engineering disciplines
Seniority
Early career / Fellow (no prior experience required, but strong technical background needed)