Research, Safety

What you'll do

• Build pre-training data filters and quality classifiers, and study their effects on safety behavior.
• Develop post-training methods, safety evaluations, and synthetic data for safety-relevant behaviors.
• Red-team models and products, identify failure modes, and design mitigations.

What they ask for

• Bachelor’s degree or equivalent experience in Computer Science, Machine Learning, Physics, Mathematics, or a related discipline, with strong theoretical and empirical grounding.
• Background in AI safety research and hands-on experience in at least one safety area, such as RLHF/RLAIF, alignment and preference modeling, deliberative alignment, safety evaluations, or red-teaming.
• Proficiency in Python and familiarity with deep learning frameworks such as PyTorch, TensorFlow, or JAX; ability to debug distributed training and write scalable code.
• Clear written communication and ability to explain complex technical concepts.

Nice to have

• Experience with long-horizon, multi-step, or agentic evaluations; synthetic data generation at scale; or modern red-teaming/jailbreaking techniques.
• AI safety research contributions, familiarity with AI safety literature and open problems, or a PhD in a related discipline or equivalent industry research experience.

Pay

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

Tools

Python, PyTorch, TensorFlow, JAX, RLHF/RLAIF

AI focus

AI safety research spanning pre-training data curation, safety-focused fine-tuning, evaluations, synthetic data, and red-teaming.