Research Engineer – Code Reinforcement Learning

About the role

Anthropic is looking for a Research Engineer to join its Code RL team. You will help advance the ability of large language models to write, edit, test, debug, and ship real software on real codebases, ensuring the process is fast, correct, and safe.

Key responsibilities

• Design and implement reinforcement‑learning environments and coding tasks for frontier models.
• Build reward signals and verification pipelines that capture what constitutes "good code".
• Run large‑scale training experiments, diagnose model performance, and iterate on improvements.
• Own end‑to‑end systems, debugging across the stack to increase speed and reliability of the research pipeline.
• Collaborate with alignment, red‑team, and production training groups to translate research breakthroughs into deployed models.

Required profile

• Strong software‑engineering background with deep Python expertise.
• Experience with asynchronous and concurrent programming.
• Ability to own complex systems end‑to‑end and troubleshoot across multiple layers.
• Comfort balancing exploratory research with robust engineering implementation.
• Passion for AI safety and building beneficial, reliable systems.

Required skills

• Python
• Async programming
• Concurrent programming
• Reinforcement learning
• RLHF (Reinforcement Learning from Human Feedback) or related post‑training techniques