Research Engineer / Scientist, Tool Use Safety

Anthropic
New York City, CA
Category Data Analyst
Job Description
Anthropic is building a team focused on making Claude the world's most capable and safe AI model. This role involves researching and implementing safe tool use and agentic application safety, focusing on the foundational layers of Claude's functionality. Collaboration with diverse research teams and deployment to production models is key to this role's success.

Requirements

  • Design and implement novel reinforcement learning methodologies.
  • Define and pursue research agendas pushing the boundaries of tool use safety.
  • Build rigorous evaluations to capture real-world safety challenges.
  • Ship research advances that protect users.
  • Collaborate with research teams and product teams.
  • Design, implement, and debug code across research and production ML stacks.
  • Strong communication and collaboration skills.

Benefits

  • Competitive compensation
  • Optional equity donation matching
  • Flexible working hours
  • Lovely office space
]]>