xAI
Model Engineer, Post-Training & RL
$180K–$600K
Palo Alto, CA · Full-time · Posted Apr 28, 2026
- Experience
- Relevant experience is not required.
About the job
Works on post-training and reinforcement learning challenges to improve AI models’ reasoning, truthfulness, and real-world capabilities.
What you'll do
- Develop reward modeling and preference optimization methods, including RLHF/DPO.
- Apply reinforcement learning to improve reasoning, truthfulness, and real-world capabilities.
What they ask for
- Production-quality coding experience in Python, PyTorch, JAX, Rust, or C++.
Nice to have
- Prior work on post-training, RLHF, or training models used by millions of people.
Pay
$180,000 - $600,000 USD (base salary)
Tools
Python, PyTorch, JAX, Rust, C++, RLHF, DPO
AI focus
Post-training and reinforcement learning, including reward modeling, preference optimization, and alignment methods, to improve model reasoning, truthfulness, and real-world capabilities.
Model-written summary · the original posting is authoritative