Research Engineer, Interpretability - Remote

Overview

Anthropic is seeking researchers and engineers to join the Interpretability team, focused on reverse engineering how trained models work to ensure AI systems are safe and beneficial.

In Short

• Join a team dedicated to mechanistic interpretability of AI models.
• Implement and analyze research experiments at scale.
• Optimize research workflows for efficiency.
• Build tools to support rapid research experimentation.
• Collaborate with teams to improve model safety.
• 5-10+ years of software building experience required.
• Proficiency in programming languages like Python, Rust, Go, or Java.
• Experience in empirical AI research is a plus.
• Strong prioritization and collaboration skills are essential.
• Competitive salary and benefits offered.

Requirements

• 5-10+ years of experience in software development.
• Proficient in at least one programming language (Python preferred).
• Experience with empirical AI research projects.
• Ability to prioritize impactful work.
• Collaborative mindset and interest in machine learning research.
• Concern for societal impacts and ethics in AI.
• Experience with large-scale distributed systems is a plus.
• Familiarity with language modeling and transformers.
• Experience with GPUs or Pytorch is beneficial.

Benefits

• Competitive compensation and benefits.
• Optional equity donation matching.
• Generous vacation and parental leave.
• Flexible working hours.
• Collaborative office space.