Research Engineer / Scientist, Alignment

Overview

Research Engineer / Scientist focused on Alignment Science, running exploratory AI safety experiments to understand and help steer risks from powerful future AI systems.

What you'll do

• Build and run machine learning experiments related to AI safety and alignment.
• Conduct exploratory experimental research on AI safety with a focus on risks from powerful future systems.
• Collaborate with teams including Interpretability, Fine-Tuning, and the Frontier Red Team.
• Train language models to subvert safety techniques and evaluate robustness.
• Run multi-agent reinforcement learning experiments such as AI Debate.
• Build tooling to evaluate the effectiveness of LLM-generated jailbreaks.
• Write scripts and prompts to generate evaluation questions for safety-relevant reasoning.
• Contribute materials to papers, blog posts, and talks.

What you'll need

• Significant software, ML, or research engineering experience.
• Experience contributing to empirical AI research projects.
• Some familiarity with technical AI safety research.
• Collaboration interest in fast-moving projects over extensive solo work.
• Care about the impacts of AI.

Details

• Interviews conducted in Python.
• Preferred candidate location: based in the Bay Area.
• Location-based hybrid policy: in offices at least 25% of the time.