Research Engineer, Reward Models

Details

Anthropic’s mission is to create reliable, interpretable, and steerable AI systems. The Reward Modeling team develops techniques for teaching AI systems to understand and embody human values and to advance AI capabilities. This role combines research and engineering to implement and integrate reward modeling advances into production systems.

Responsibilities

• Implement novel reward modeling architectures and techniques
• Optimize model training pipelines
• Build and optimize data pipelines
• Collaborate across teams to integrate reward modeling advances into production systems
• Communicate engineering progress through internal documentation and potential publications

Requirements

• Strong engineering background in machine learning
• Demonstrable expertise in preference learning, reinforcement learning, deep learning, or related areas
• Proficiency in Python (required)
• Experience with deep learning frameworks and distributed computing
• Familiarity with modern LLM architectures and alignment techniques
• Experience improving model training pipelines and building data pipelines
• Comfortable with experimental frontier AI research; view research and engineering as complementary
• Ability to clearly communicate complex technical concepts and research findings
• Deep interest in AI alignment and safety
• Education: at least a Bachelor's degree in a related field or equivalent experience

Notes:

• Experience with reward models is not required; experience with LLMs or other large models is a significant plus.
• Anthropic welcomes candidates at various experience levels, with a preference for senior engineers who have hands-on experience with frontier AI systems.

Logistics

• Locations: San Francisco, CA; New York City, NY; Seattle, WA
• Location-based hybrid policy: staff expected to be in office at least 25% of the time (some roles may require more)
• Visa sponsorship: Anthropic does sponsor visas and retains an immigration lawyer to assist; not all roles/candidates can be successfully sponsored

Benefits

• Competitive compensation (see salary range)
• Optional equity donation matching
• Generous vacation and parental leave
• Flexible working hours
• Office space for collaboration

How we're different

Anthropic pursues large-scale, high-impact AI research as a cohesive team, emphasizing communication and collaboration. The team values empirical, interdisciplinary approaches and publishes and shares research directions related to scaling laws, interpretability, learning from human preferences, and AI safety.