San Francisco paper
Tech and design

Research engineer, reinforcement learning

AnthropicSan Francisco, New York and Seattle

As printed
Research engineer, RL $500–850k. Anthropic. Hybrid.Ref. 7912
  • Pay$500,000–850,000 per year
  • WhereSan Francisco, New York and Seattle
  • WorkHybrid
  • TermsFull-time
  • ReferenceRef. 7912
  • In the paper untilThu 19 Nov

Skills and tools

  • Reinforcement learning
  • Transformer architecture
  • Distributed training
  • Data parallelism
  • Tensor parallelism
  • Pipeline parallelism
  • Python
  • JAX
  • PyTorch
  • Scaling laws

You’ll need

  • Bachelor's degree

Perks

  • Equity donation matching
  • Paid time off
  • Parental leave
  • Flexible hours

About the job

Hybrid, at least 25% in an office. For candidates with experience training large models in distributed settings and strong Python, JAX, or PyTorch skills. Bachelor's degree or equivalent experience. Flexible hours, paid time off, parental leave, and optional equity donation matching.

This is our short summary. Read the full posting on boards.greenhouse.io.

Found on boards.greenhouse.io. Not placed by the employer; applications go to the original posting.

Is this your job? Claim it
Report this ad
What’s wrong?
Apply on boards.greenhouse.io