Mercor Research Fellowship — APEX

📋 Description

• Propose and scope a new benchmark or evaluation technique in a domain APEX doesn’t yet cover well
• Design task specifications and grading rubrics with Mercor’s network of domain experts.
• Build and validate the benchmark: pilot tasks, calibrate scoring, and test for contamination.
• Run frontier models against your benchmark and analyze failures.
• Publish results as papers, datasets, leaderboards, or methodologies for internal adoption.
• Collaborate with Mercor’s research and engineering teams to integrate learnings into APEX.

🎯 Requirements

• Genuine interest in evaluation as a research discipline.
• Background in CS, ML, statistics, or adjacent fields; no ML publication requirement.
• A specific, well-scoped idea for a benchmark or eval technique.
• Comfortable in a startup environment with fast iteration.
• Commit at least 20 hours/week for the duration of the fellowship.
• Bonus: experience with agentic evaluation, RL environments, or domain expertise in

🎁 Benefits

• 3 month stipend of $40,000 or 6 month stipend of $80,000.
• Unlimited API credits, GPU compute budget, and paid expert/human-data time.
• Weekly mentorship with the APEX research team and access to broader research org.
• Access to frontier model APIs and enterprise evaluation problems.
• Optional desk in Mercor’s San Francisco office; introductions to researchers across frontier labs
• High-potential fellows may receive a full-time offer after the fellowship.