Research Engineer / Scientist - Reinforcement Learning

Role overview

Develop reinforcement learning methods and training systems for open models, with work shaped around your strengths in algorithms, scaling research, or training frameworks. The role connects research to large-scale implementation, supporting capabilities such as reasoning, coding, and long-horizon task completion.

Responsibilities - Research and test RL approaches, including training objectives and policy optimization methods. - Investigate instability in asynchronous training, such as stale policies, mismatches between trainers and samplers, and numerical differences. - Build and improve training frameworks and systems that support reliable, scalable experiments. - Analyze training runs, diagnose failures, and share findings with research and engineering collaborators. - Contribute to work on coding agents, verifiable rewards, or environments for extended tasks, depending on your focus.

Requirements

The available source describes the role’s research areas and examples of relevant work, but does not provide enough detail to state specific qualifications.

Benefits and work setup - Full-time role with remote work or in-person work in San Francisco. - Stated cash compensation range of $150,000–$350,000, including equity incentives. - Visa sponsorship and relocation assistance are available for international candidates. - Quarterly team off-sites, hackathons, conferences, and learning opportunities.