Frontier-Scale RL Research Engineer

EngineersOfAI; New York (NY); On-site; USD 180,000 - 240,000; Full time

Job summary

Anthropic seeks an ML researcher to bridge research and engineering, focused on RL Scaling. You will design and implement next-generation RL architectures and algorithms, pushing frontier-scale runs and diagnosing why results differ as scale grows.

You’ll shape how fast teams iterate and how efficiently experiments run. You will also build the systems that enable rapid experimentation, own end-to-end performance of the largest RL runs, and contribute to models that quantify cost and speed

Qualifications

• Deep familiarity with modern transformer language models, training dynamics and large-scale optimization.
• Hands-on experience training large models in a distributed setting with data, tensor, and pipeline parallelism.
• Track record of original ML research or systems work with demonstrated impact.
• Ability to design rigorous scale experiments with baselines and ablations.
• Ability to reason quantitatively about compute, memory and communication costs.
• Strong coding skills in Python and JAX or PyTorch across the full stack.

Responsibilities

• Study how RL training and sampling scale with model size, context length, and compute, and find algorithmic and systems changes that keep scaling efficient.
• Develop next-generation model architectures and RL algorithms, and run them efficiently at frontier scale.
• Take promising small-scale results to frontier-scale runs and diagnose causes across numerical, algorithmic, or systemic dimensions.
• Build the experimental infrastructure for fast, reproducible comparisons of architecture and algorithm variants at meaningful scale.
• Own end-to-end performance of our largest RL runs, from research code to hardware.
• Build performance and cost models to guide architecture and algorithm choices at scale.
• Investigate training dynamics at scale, including instabilities and throughput regressions, and trace root causes.

Skills

Transformer models

Python

JAX

PyTorch

Distributed training

Experiment design

Tools

C++

Rust

Job description

Anthropic seeks an ML researcher to bridge research and engineering, focused on RL Scaling. You will design and implement next-generation RL architectures and algorithms, pushing frontier-scale runs and diagnosing why results differ as scale grows.

You’ll shape how fast teams iterate and how efficiently experiments run. You will also build the systems that enable rapid experimentation, own end-to-end performance of the largest RL runs, and contribute to models that quantify cost and speed