Research Engineer, AI Safety & Alignment

What you'll do

• Develop safety and alignment evaluation methods and metrics, and research alignment, value learning, and interpretability techniques.
• Adversarially test models and mitigate bias, toxicity, and other harmful behaviors using methods such as RLHF and fine-tuning.
• Work with engineering and product teams to put safety research into practice; contribute to the research community through publications and presentations.

What they ask for

• PhD (or equivalent experience) in a relevant field such as Computer Science or Machine Learning.
• Write clear, clean production-facing and training code.
• Experience working with GPUs for training, serving, and debugging.
• Experience with data pipelines and data infrastructure; strong understanding of modern machine learning, particularly transformers and reinforcement learning.

Nice to have

• Experience with product experimentation and A/B testing or training large models in a distributed setting.
• Familiarity with ML deployment and orchestration, explainable AI, or AI safety research.

Pay

$225K – $400K USD base salary

Tools

GPUs, transformers, reinforcement learning, RLHF, Kubernetes, Docker

AI focus

AI safety and alignment research, including model evaluation, adversarial testing, value learning, interpretability, and mitigation of harmful model behavior.