Research, Post-Training Evals

What you'll do

• Create evaluations and research signals for model capabilities and behaviors; assess usefulness in research and product workflows.
• Improve evaluation robustness and develop benchmark auditing methodologies.
• Build agentic evaluation environments and user simulators, and evaluate preferences, personalization, biases, and values.

What they ask for

• Bachelor’s degree or equivalent experience in a listed or related discipline, with strong theoretical and empirical grounding.
• Experience designing, building, or analyzing evaluations, benchmarks, datasets, graders, or other measurement systems.
• Strong written and verbal communication and ability to collaborate across research and engineering teams.

Nice to have

• Experience with LLMs, post-training, reinforcement learning, agentic systems, evaluation auditing, human evaluations, LLM-judges, or open-ended task evaluation.
• Proficiency in Python and familiarity with a deep learning framework; experience with evaluation methods, research signals, or nuanced model behaviors.

Pay

$350,000 - $475,000 USD annual salary (base salary)

Tools

Python, PyTorch, TensorFlow, JAX

AI focus

Develops and audits evaluations for model capabilities, behaviors, usability, and post-training; includes specialized agentic evaluation environments and user simulators.