Research Scientist — Alignment

About the Role

Mistral AI is hiring a Research Scientist to work on alignment techniques for open-weight language models. You will develop methods for steering model behavior, design evaluation frameworks, and contribute to making powerful open models safer.

Requirements

• PhD or equivalent in ML with focus on alignment or safety
• Research experience with RLHF, DPO, or constitutional methods
• Strong implementation skills in Python and PyTorch
• Publications in AI safety or alignment venues
• Deep understanding of LLM behavior and failure modes

Nice to Have

• Experience with open-weight model alignment
• Background in reward modeling
• Familiarity with European AI regulation
• Multilingual evaluation experience

Benefits

Skills

Related Jobs