Researcher, Alignment Science

OpenAI

San Francisco, California, United States
Not specified; not specified; relocation assistanc...
3d onsite
Hands-on experience training large ml models
Strong python and pytorch engineering skills
Experience with reinforcement learning and post-training
The Alignment Science team focuses on scalable methods to ensure instruction-following, honesty, and robustness as models become more capable

Job Summary

  • The Alignment Science team focuses on scalable methods to ensure instruction-following, honesty, and robustness as models become more capable.
  • You will design and run experiments to help models follow user intent, remain calibrated about correctness, and honestly surface their own mistakes.
  • The role offers a hybrid work model requiring 3 days in the office per week in San Francisco, with relocation assistance available.

Matching Summary

The Alignment Science team focuses on scalable methods to ensure instruction-following, honesty, and robustness as models become more capable.

Salary

Not specified; Not specified; Relocation assistance offered

Skills & Requirements

Must-have

  • Hands-on experience training large ML models
  • Strong Python and PyTorch engineering skills
  • Experience with reinforcement learning and post-training
  • Ability to design quantitative research experiments

Nice-to-have

  • Competitive programming or math contest background
  • Experience with scalable oversight methods
  • Track record of publishing external research
  • Comfort operating independently in fast-paced environments

Key Requirements

  • Strong hands-on experience with LLMs
  • Mathematical rigor and quantitative taste
  • High independence and self-direction

Work Rights

Not specified

Tailored Resume

Cover Letter