Cursor
Research Scientist
San Francisco, CA +1 · On-site · Full-time · Posted Jan 27, 2026
- Location
- San Francisco, CA
New York, NY
About the job
Research Scientist developing reinforcement learning and mid-training methods for frontier coding agents, from research hypotheses and experiments through training, evaluation, and model improvements.
What you'll do
- Research RL methods for longer-horizon tasks and more compute-efficient training.
- Train graders for coding tasks with non-verifiable rewards and improve the quality and difficulty of training data.
- Develop real-time reinforcement learning for coding agents.
What they ask for
- Deep background in RL and strong machine learning fundamentals.
- Excellent programming and software engineering skills.
- Ability to handle ambiguous research tasks with little guidance.
- Care for data quality and willingness to investigate data.
AI focus
Research and train reinforcement-learning and mid-training methods for coding agents, including graders, training data, and real-time RL.
Model-written summary · the original posting is authoritative