Research Scientist Multi Agent
AI safety and research company building reliable, interpretable, and steerable AI systems, including the Claude product family and developer platform.
Maintainer signals as of 9/25/2026
Funding history
About Anthropic
Anthropic PBC develops frontier AI systems and deploys them through Claude products and the Claude Platform, with a stated focus on safety, interpretability, and steerability.
Skills
Candidate Availability
Required and preferred rules are kept separate and reflect the wording in the original posting.
About the Role
You will create training environments and data for agentic tasks, experiment with agent harness configurations, and build quantitative benchmarks. You will evaluate multi-agent performance at scale and collaborate with product stakeholders to solve difficult applications of agents.
Requirements
- Experience with large-scale reinforcement learning on language models
- Experience training multi-agent systems
- Knowledge of incentives and mechanism design
- Communication skills
Responsibilities
- Create and optimize model-training environments and data for agentic tasks
- Develop and compare agent harness configurations, including memory, context management, and communication architectures
- Design and implement quantitative benchmarks for large-scale agentic tasks
- Partner with product stakeholders to solve challenges in applying agents to products
- Design reinforcement-learning environments for groups of agents
- Build agent affordances and evaluations for collaborative agent behavior
Benefits
- Optional equity donation matching
- Generous vacation
- Parental leave
- Flexible working hours
- Visa sponsorship support