Receive alerts when this company posts new jobs.

Similar Jobs

Job Details

Research Scientist, Clarity

at OpenAI

Posted: 9/14/2020
Job Status: Full Time
Job Reference #: c66b49a8-56ba-4af1-9aa5-04f7b34f7b8d
Keywords: research, scientist

Job Description

Research Scientist, Clarity

San Francisco /
Clarity /

Apply for this job

One of the greatest mysteries of neural networks is how they can accomplish things no human knows how to write a computer program to solve. The OpenAI Clarity team aims to reverse engineer these systems, turning them back into human-understandable computer programs, and answer this mystery. We do this by studying which features exist in the model, and how weights and features combine together into
circuits implementing meaningful algorithms

As a research scientist on the Clarity team, you will work towards cracking this mystery and understanding the inner workings of neural networks, from standard open source models, to giant models like GPT-3 that OpenAI has built.

Some of the major questions you may think about include:

What features exist in the hidden layers of various models, enabling them to accomplish the tasks they do?

How are these features built from previous features? What algorithms are embedded in the weights of the network?
How is knowledge stored in neural networks? What kinds of reasoning can they do?
How can we scale this form of analysis to truly gigantic models?
How can our understanding of these models help mitigate bias and safety concerns, or enable more humane ways of interacting with AI systems?

You will also:

Contribute to infrastructure needed to answer these questions.

We are looking for you to:

Have a strong research background in machine learning, or a related technical field involving reverse engineering complex systems (neuroscience, systems biology, software reverse engineering, etc), demonstrated in publications or projects.

Have a strong understanding of fundamental concepts in deep learning and related mathematics.
Have strong software engineering skills and feel comfortable diving into engineering challenges as necessary for the research problems you are pursuing.
Possess the ability to own and pursue a research agenda, including choosing impactful research problems aligned with the teams goals and autonomously carrying out long-running projects.
The ability to quickly implement research ideas, iterate on them, and evaluate them empirically.
Be aligned with the circuits agenda (see above) as a path to reverse engineering neural networks (or come to us with a proposal of another path and win us over!).

We also value:

Having done projects related to our research. Our flavor of interpretability research is uncommon, but we’d be very excited to see candidates who have done a related project.

Interest in data visualization, interfaces, and experience using web technologies to build them. We believe that the ability to build visualizations and interfaces is a huge enabler in studying neural networks.

About OpenAI

We’re building safe Artificial General Intelligence (AGI), and ensuring it leads to a good outcome for humans. We believe that unreasonably great results are best delivered by a highly creative group working in concert. We are an equal opportunity employer and value diversity at our company. We do not discriminate on the basis of race, religion, color, national origin, gender, sexual orientation, age, marital status, veteran status, or disability status.

This position is subject to a background check for any convictions directly related to its duties and responsibilities. Only job-related convictions will be considered and will not automatically disqualify the candidate. Pursuant to the San Francisco Fair Chance Ordinance, we will consider for employment qualified applicants with arrest and conviction records.


- Health, dental, and vision insurance for you and your family
- Unlimited time off (we encourage 4+ weeks per year)
- Parental leave
- Flexible work hours
- Lunch and dinner each day
- 401(k) plan

Apply for this job