Elicit
ML Research Resident
Oakland, CA (or remote within US timezones). Posted 23 months ago
The role
You will work for 3 months developing computational procedures that improve a knowledge state over thousands of iterations. The work focuses on designing and testing improvement operators that maintain stability while making genuine progress.
What you will do
- Design and test improvement operators that maintain stability over 1000+ iterations
- Start with simple cases like shallow refactoring of scientific papers
- Demonstrate reliable iteration before scaling to more complex reasoning tasks
Required qualifications
- Experience with LLMs
- Good intuitions about what makes reasoning systematic and verifiable
- Care about AI transparency
Nice to have
- Strong software engineering background
- Concrete examples of how you've applied software engineering background to come up with novel abstractions that push the frontiers of automated reasoning
Benefits
- Potential of full-time offer for exceptional candidates
- Quarterly retreat
About the company
Elicit is building a research agent that can use an unlimited amount of test-time compute while keeping its reasoning transparent and verifiable.