Research Engineer

Core responsibilities:

• Develop scalable implementations of machine learning algorithms and use them to run scientific experiments
• Contribute to open source codebases such as PyTorch, HuggingFace Transformers and Accelerate
• Work alongside scientists on write-up of results and be credited as an author in papers
• Propose novel research directions and join research pods based on interests

This role contributes to ensuring advanced AI is safe and beneficial for everyone by conducting breakthrough AI safety research, advancing global understanding of AI risks and solutions, and fostering a coordinated global response.

Required qualifications:

• Significant software engineering experience or experience applying machine learning methods (evidenced by prior work, open-source contributions, or publications)
• Experience with at least one object-oriented programming language (preferably Python)
• Results-oriented and motivated by impactful research

Preferred qualifications:

• Experience with ML frameworks like PyTorch or TensorFlow
• Experience with natural language processing or reinforcement learning
• Knowledge of operating system internals and distributed systems
• Publications or open-source software contributions
• Basic linear algebra, calculus, probability, and statistics

Compensation: $100,000-$190,000/year depending on experience and location

• Work-related travel and equipment expenses covered
• Catered lunch and dinner at Berkeley offices
• Time to study and attend conferences free of charge
• Engineering mentorship via code review, pair programming and regular 1-to-1s

FAR.AI is a non-profit AI research institute dedicated to ensuring advanced AI is safe and beneficial for everyone. Their mission is to facilitate breakthrough AI safety research, advance global understanding of AI risks and solutions, and foster a coordinated global response through three prongs: FAR.Research (conducting cutting-edge AI safety research), FAR.Futures (bringing together key stakeholders to drive change), and FAR.Labs (hosting a co-working space to incubate other AI safety organizations).