500K–850K USD / year
Research Engineer / Research Scientist, RL Frontiers
Machine learningHybrid — San Francisco, CA | New York City, NY | Seattle, WA
Published on 2026-09-29
These details were extracted automatically from the original listing. They may not be complete or fully up to date — it's worth checking the original job post.
About this role
Join the RL Scaling team to develop and scale reinforcement learning algorithms and architectures. You'll work on enhancing model performance and efficiency, running experiments, and resolving challenges encountered at scale. This role requires a mix of research and engineering skills in AI systems, particularly in reinforcement learning.
About the company
Anthropic builds reliable and interpretable AI systems, including the Claude family of models, for businesses and developers. The company focuses on AI safety research and practical AI assistants for enterprise and consumer use.
Stack
PythonJAXPyTorch
What you'll do
- Study RL training and sampling scaling
- Develop next-generation RL algorithms and architectures
- Diagnose performance differences at scale
- Build experimental infrastructure for rapid research
- Own end-to-end performance of large RL runs
- Build performance and cost models
- Investigate training dynamics at scale
What we're looking for
- Familiarity with transformer language models
- Experience with large model training in distributed settings
- Technical work in ML training or systems
- Designing experiments at scale
- Quantitative reasoning about model costs
- Programming skills in Python and relevant frameworks
Nice to have
- Research experience in reinforcement learning
- Experience with scaling laws
- Designing transformer architectures
- Debugging large-scale training issues
- Understanding of numerics in training
- Experience with C++ or Rust