Research Scientist, Safety Post Training
Scale AI · San Francisco, CA; New York, NY · San Francisco, CA
CHECKED 14H AGO
Infrastructureonsite
At a glance
Highlights
- Study how training choices affect safety and alignment of frontier models.
- Build interpretability evaluations revealing why models behave unsafely.
- Publish safety research guiding governments and industry on AI risk.
What they offer
- Base salary with equity grant
- Comprehensive health, dental and vision coverage
- Retirement benefits
- Learning and development stipend
- Generous PTO
Ideal candidate
Mission-driven ML researcher passionate about safe AI who enjoys prototyping post-training ideas and translating findings into real-world standards with policymakers and cross-functional partners.
Requirements
- Experience with post-training and RL techniques such as RLHF, DPO, GRPO and similar approaches
- Track record of published research in machine learning, particularly in generative AI
- At least three years of experience addressing sophisticated ML problems in research or product development
- Strong written and verbal communication skills to operate in a cross-functional team
- hiring process
- Research interviews assessing practical ML prototyping and debugging, research concepts and culture fit; no LeetCode-style questions.
Lab profile
About Scale AI
AI data foundry providing high-quality data for training, fine-tuning, and evaluating AI systems.
What they build
- Training, fine-tuning, and evaluation data for frontier models
- Multimodal and coding data for audio, image, video and world models
- Custom AI applications for enterprise and global public sector
Known for
- Leading data partner for frontier model developers
- Trusted across enterprise, defense and public sector
- Depth in multimodal, coding and physical AI
- End-to-end capability from strategy to delivery and evaluation
Key facts
- Founded in 2016
- Headquartered in San Francisco, California
- Private company backed by Accel, Founders Fund, Index Ventures, Nvidia and Meta