Researcher, Interpretability

Core responsibilities:

• Perform cutting-edge research with large language and vision models to improve understanding of deep neural network features
• Plan and run interpretability experiments, analyzing the results
• Implement and maintain open source interpretability tools, such as concept erasure
• Review and synthesize academic literature relevant to the interpretability of AI systems
• Communicate research findings to the wider AI community through various channels

This role advances AI safety by improving understanding of deep neural networks through interpretability research, helping to make AI systems more transparent and trustworthy.

Required skills and qualifications:

• Experience with large language and vision models
• Ability to plan and run interpretability experiments
• Proficiency in implementing and maintaining open source interpretability tools
• Strong academic literature review and synthesis skills
• Excellent communication skills for sharing research findings

No benefits information provided in the job description.

EleutherAI is an AI research organization focused on open source AI development and interpretability research.