Researcher, Synthetic RL

The Synthetic RL team develops reinforcement learning methods that leverage synthetic data, environments, and feedback to train and evaluate frontier AI models. The team explores approaches such as self-play, simulators, and other synthetic evaluations to push model capability, generalization, and alignment beyond what is possible with the current prevailing methodology.

About the Role

As a Research Scientist on the Synthetic RL team, you will develop novel reinforcement learning techniques that use synthetic environments and feedback to improve large-scale models. You’ll work closely with other researchers to design experiments, analyze learning dynamics, and translate research insights into training approaches used in production systems.

We’re looking for researchers who enjoy working on open-ended problems, value fast iteration, and want their work to directly shape how frontier models are trained.

This role is based in San Francisco, CA. We use a hybrid work model of 3 days in the office per week and offer relocation assistance to new employees.

In this role, you will:

• Research and develop reinforcement learning algorithms

• Design and run experiments to study training dynamics and model behavior at scale

• Collaborate with engineers and researchers to integrate successful approaches into model training pipelines
  

• Have a strong background in reinforcement learning, machine learning research, or related fields

• Have strong engineering and statistical analysis skills

• Enjoy exploring new problem spaces where data, objectives, and evaluation are imperfect or evolving

• Are motivated by seeing research ideas influence real-world AI systems