New Model Behavior Architect, Safety

Mistral AI Mistral AI · AI Frontier · Paris, France · Research

Mistral AI is seeking a Model Behavior Architect for their Safety team to define and measure LLM behavior across reasoning, audio, alignment, and tool use. The role involves interacting with models, gathering feedback, designing evaluations, and collaborating with AI Scientists to ensure models meet high standards of performance and reliability.

What you'd actually do

  1. Interact with models to pinpoint areas where behaviour can be refined and optimised.
  2. Gather feedback from internal and external sources to identify and prioritise improvement opportunities.
  3. Design and implement evaluations, data guidelines, and synthetic testing environments.
  4. Address edge cases through rigorous testing to enhance model reliability and accuracy.
  5. Develop robust evaluation pipelines to assess the capabilities of model candidates.

Skills

Required

  • Deep understanding of linguistics, language, and translation, or expertise in engineering and code behaviour, or experience with LLM agents, including reasoning and tool use.
  • Prior knowledge in training and optimising model behavior for real-world applications.
  • Expertise in building robust, scalable evaluation frameworks for AI systems.
  • Ability to thrive in dynamic, technically complex environments and deliver innovative solutions.
  • Track record of solving open-ended challenges with creative, practical approaches.

What the JD emphasized

  • rigorous testing
  • robust evaluation frameworks
  • open-ended challenges

Other signals

  • model behavior
  • evaluation frameworks
  • safety
  • alignment