Research Engineer, Code RL (Reinforcement Learning)

About the role

We're hiring for the Code RL team within the RL organization. As a Research Engineer, you'll advance our models' ability to write, edit, test, debug, and ship real software — end to end, on real codebases, with real tools — and to do it correctly, fast, and safely. This role blends research and engineering. You'll design RL environments and coding tasks, build the reward signals and verifiers that capture what "good code" means, run training experiments on frontier models, diagnose why a model does (or doesn't) get better at a class of software-engineering work, and improve the speed and reliability of the pipelines that make all of that iterate fast.

Responsibilities

• Advance our models' ability to write, edit, test, debug, and ship real software — end to end, on real codebases, with real tools — and to do it correctly, fast, and safely.
• Design RL environments and coding tasks.
• Build the reward signals and verifiers that capture what "good code" means.
• Run training experiments on frontier models.
• Diagnose why a model does (or doesn't) get better at a class of software-engineering work.
• Improve the speed and reliability of the pipelines that make all of that iterate fast.

Requirements

• Strong software-engineering skills and deep Python expertise, including async/concurrent programming.
• Comfortable owning systems end to end and debugging across the stack.
• Career-long experience balancing research exploration with engineering implementation, and engaging rigorously in shaping experimental design and interpreting results.
• A passion for code quality, testing, and performance.
• A commitment to developing safe and beneficial AI systems.

Qualifications

• Bachelor’s degree or an equivalent combination of education, training, and/or experience in a field relevant to the role as demonstrated through coursework, training, or professional experience.

Skills

• Experience with reinforcement learning, RLHF, post-training, or LLM finetuning.
• Experience with coding agents, code-execution sandboxes, eval 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.

Benefits

Annual compensation range for this role is $500,000—$850,000 USD.

Pay

The annual compensation range for this role is listed above.

Schedule

Currently, we expect all staff to be in one of our offices at least 25% of the time. However, some roles may require more time in our offices.

Logistics

• Minimum education: Bachelor’s degree or an equivalent combination of education, training, and/or experience.
• Required field of study: A field relevant to the role as demonstrated through coursework, training, or professional experience.
• Minimum years of experience: Years of experience required will correlate with the internal job level requirements for the position.

Location-based hybrid policy

Location-based hybrid policy: Currently, we expect all staff to be in one of our offices at least 25% of the time. However, some roles may require more time in our offices.

Visa sponsorship

We do sponsor visas! However, we aren't able to successfully sponsor visas for every role and every candidate. But if we make you an offer, we will make every reasonable effort to get you a visa, and we retain an immigration lawyer to help with this.

How We're Different

• We believe that the highest-impact AI research will be big science. At Anthropic we work as a single cohesive team on just a few large-scale research efforts.
• We value impact — advancing our long-term goals of steerable, trustworthy AI — rather than work on smaller and more specific puzzles.
• We view AI research as an empirical science, which has as much in common with physics and biology as with traditional efforts in computer science.
• We're an extremely collaborative group, and we host frequent research discussions to ensure that we are pursuing the highest-impact work at any given time.
• The easiest way to understand our research directions is to read our recent research. This research continues many of the directions our team worked on prior to Anthropic, including: GPT-3, Circuit-Based Interpretability, Multimodal Neurons, Scaling Laws, AI & Compute, Concrete Problems in AI Safety, and Learning from Human Preferences.

Guidance on Candidates' AI Usage

Learn about our policy for using AI in our application process.