ML Research Resident

About the role

Elicit is building a research agent that uses unlimited test-time compute to improve knowledge states over thousands of iterations. The role involves working on computational procedures (operators) that enhance structured information like scientific papers.

Responsibilities

• Design and test improvement operators that maintain stability over 1000+ iterations while making genuine progress.
• Start with simple cases (e.g., shallow refactoring of scientific papers) and scale to more complex reasoning tasks.
• Develop systems that address core challenges in AI transparency and scalable reasoning.

Requirements

• Experience with LLMs.
• Good intuitions about systematic reasoning and verifiability.
• Care about AI transparency.
• A strong software engineering background.
• Concrete examples of applying this background to novel abstractions in automated reasoning.

Qualifications

• Strong academic background in computer science or related fields.
• Experience with natural language processing and machine learning.
• Understanding of knowledge representation and reasoning.

Skills

• Programming skills (Python, Java, etc.).
• Knowledge of LLMs and their limitations.
• Experience with iterative reasoning and verification.

Benefits

• In-person or remote work location in the US.
• Potential for a full-time offer for exceptional candidates.
• Quarterly retreats at fun locations for the whole team.

Pay

$12-15k/month depending on experience.

Schedule

3-month contract role.