Job Description

About the role:

You’re excited to work at an organization that functions as a single, cohesive team pursuing large-scale AI research projects. You’re working to align state of the art models with human values and preferences, understand and interpret deep neural networks, or develop new models to support these areas of research. You view research and engineering as two sides of the same coin, and seek to understand all aspects of our research program as well as possible, to maximize the impact of your insights. You have ambitious goals for AI safety and general progress in the next few years, and you’re working to create the best outcomes over the long-term.

Note: This is an "evergreen" role that we keep open on an ongoing basis. We receive many applications for this position, and you may not hear back from us directly if we do not currently have an open role on any of our teams that matches your skills and experience. We encourage you to apply despite this, as we are continually evaluating for top talent to join our team. You are also welcome to reapply as you gain more experience, but we suggest only reapplying once per year.

We may also put up separate, team-specific job postings. In those cases, the teams will give preference to candidates who apply to the team-specific postings, so if you are interested in a specific team please make sure to check for team-specific job postings!

You may be a good fit if you:

1. (Required) Have at least 1 research project related to machine learning in which you played a major role.

Strong candidates may also:

• Clearly articulate and discuss the motivations behind your work, and teach us about what you've learned.
• Like writing up and communicating your results, even when they're null.
• Like working on a team and making collaborative, data-driven research decisions.
• Believe that a great way to discover a new, big-picture vision is to first get involved in the details.
• Have experience doing quantitative science in another field.

Representative projects:

• Design a new method for automatically red teaming language models for harmful behavior.
• Create evaluation datasets for potential, new dangerous capabilities of language models such as persuasion or deception.
• Compare the compute efficiency of two Transformer variants.
• Create an interactive visualization of attention between tokens in a language model.