Research Scientist - RL Training
Snorkel AI · New York City, NY (Hybrid) · San Francisco, CA (Hybrid) · United States (Remote) · Remote
Infrastructurehybrid$200k–$375k
At a glance
Highlights
- Advance RL training to help LLMs solve frontier-level, long-horizon tasks.
- Extend RL to complex multi-turn settings with tool use and varied feedback.
- Drive research direction connecting training results to better tasks and data.
Ideal candidate
RL researcher passionate about pushing LLM reasoning on hard, long-horizon problems. You love diagnosing training, closing the loop with data, and turning insights into published advances.
Requirements
- Deep expertise in reinforcement learning for language models, including policy optimization, credit assignment, and learning from sparse or delayed rewards
- Experience designing or adapting methods for multi-turn and long-horizon tasks
- Experience training or fine-tuning 30B+ large language models at scale with distributed training infrastructure
- Strong proficiency in Python and ML frameworks, especially PyTorch and HuggingFace, and hands-on experience with RL frameworks such as Verl
- Strong software engineering skills building research systems and training pipelines others can run, extend, and maintain
- Familiarity with ML infrastructure and cloud platforms and tools such as AWS, GCP, Kubernetes, Slurm
- Ability to analyze training behavior and evaluation results
- Ph.D. in machine learning, reinforcement learning, or related field strongly preferred; exceptional industry experience considered
Lab profile
About Snorkel AI
Frontier AI data lab building datasets, benchmarks, and evals for frontier and agentic AI.
What they build
- Datasets for frontier AI
- Benchmarks and evals
- Custom AI data solutions
Known for
- Research-driven AI data development
- Combines technology with expert data development
- Works with leading AI labs and enterprises
Key facts
- Founded in 2019
- Spun out of Stanford AI Lab
Posted · Last seen