Agent Post-Training, Frontier Evals and Environments Research

Researcher building frontier RL environments, evaluations, and training signals to steer OpenAI's largest agent training runs and measure model capabilities.

About the job

Responsibilities

Create ambitious RL environments to push frontier models to their limits and measure model capabilities, skills, and behaviors
Develop new methodologies for automatically exploring model behavior
Dive deep into the science of measurement, including scalability, reliability, and variance of evaluation methodology
Help steer training for the largest training runs
Design scalable systems and processes to support continuous evaluation
Build self-improvement loops to automate model understanding

Requirements

Strong technical fundamentals in machine learning, software engineering, systems, statistics, or a related field
Hands-on experience with LLMs, RL, RLHF/RLAIF, post-training, evals, graders, synthetic data, model training, coding agents, tool-using agents, or production ML systems
Ability to move from a vague behavioral problem to a concrete experiment: define the hypothesis, build the pipeline, run the model, analyze the result, and decide next steps
Comfortable working across research, product, infrastructure, data, evals, and safety boundaries

Nice-to-Haves

Excitement for open-ended problems where the path is unclear and the signal is noisy
Care about product impact and model behavior beyond benchmark movement
Opinions about what makes an agent useful, reliable, honest, tasteful, and easy to work with
Willingness to build load-bearing systems and processes even when the work is not glamorous

Skills

Machine Learning, Software Engineering, Statistics, LLMs, Reinforcement Learning, RLHF, Rlaif, Post-Training, Evaluations, Graders, Synthetic Data, Model Training, Coding Agents, Tool-Using Agents, Production Ml Systems