This role is for a Research Engineer on the Code RL team at Anthropic, focusing on developing reinforcement learning systems that enable AI models to write, edit, test, debug, and ship software. The position blends research and engineering, requiring strong software‑engineering skills, deep Python expertise, and experience with reinforcement learning and large‑scale ML systems. The work is hybrid, based in San Francisco with occasional presence in New York City, and offers a competitive salary range of $500,000–$850,000 USD, along with benefits such as equity donation matching, vacation, parental leave, and visa sponsorship.
Research Engineer, Code RL (Reinforcement Learning) at Anthropic
Hybrid - San Francisco, CA | New York City, NY
More jobs at AnthropicSalary
USD 500,000 - 850,000
Requirements
Skills
- Bachelor’s degree in a field relevant to the role or equivalent experience
- Strong software‑engineering skills and deep Python expertise, including async/concurrent programming
- Comfortable owning systems end to end and debugging across the stack
- Ability to balance research exploration with engineering implementation and rigorously shape experimental design
- Care for code quality, testing, and performance
- Passion for the potential impact of AI and commitment to safe and beneficial systems
- Experience with reinforcement learning, RLHF, post‑training, or LLM fine‑tuning
- Built coding agents, code‑execution sandboxes, evaluation harnesses, verifiers, or developer tooling
- Background in program analysis, testing, verification, compilers, or formal methods
- Experience with PyTorch and large‑scale distributed training; performance profiling and optimization of ML systems
- CUDA / GPU or TPU kernel experience and accelerator‑performance intuition
- Experience with virtualization and sandboxed code execution environments
Responsibilities
- Design RL environments and coding tasks
- Build reward signals and verifiers that capture what "good code" means
- Run training experiments on frontier models
- Diagnose model behavior and improve speed and reliability of pipelines
- Iterate fast to enhance agentic coding behaviors, code correctness, long‑horizon autonomous engineering, and high‑performance code for accelerators
Technologies
PythonPyTorchReinforcement learningRLHFLLM fine‑tuningCUDAGPUTPUPerformance profilingVirtualizationSandboxed code executionLarge‑scale distributed training
See if your resume is ready for this job
See how our AI can optimize your resume and improve your chances for this role.