← All Jobs
Posted Apr 16, 2026

Research Expert

Apply Now
**Role Overview** Mercor is collaborating with a leading AI research organization to engage experts capable of developing challenging Q&A pairs derived from academic papers on arXiv. This short-term project aims to enhance benchmark datasets by crafting questions designed to surpass the capabilities of state-of-the-art AI models. Experts will leverage their deep knowledge of specialized research domains to create analytically demanding and insightful prompts for evaluating advanced AI systems. • *Key Responsibilities** - Select and thoroughly analyze academic papers from arXiv in your area of expertise. - Craft sophisticated Q&A pairs that test advanced AI models’ conceptual, analytical, and synthesis abilities. - Clearly document detailed metadata, including paper references, question categories, difficulty justifications, and task completion times. - Evaluate model-generated answers, iteratively refining questions until leading AI models fail. • *Ideal Qualifications** - Advanced degree (PhD or equivalent) or significant professional experience in a specialized research field (e.g., computational biology, quantum physics, deep learning, etc.). - Proven familiarity with academic literature and research publication standards (e.g., arXiv). - Strong analytical reasoning, meticulous attention to detail, and clarity in technical writing. - Prior experience with benchmark design or AI evaluation preferred. - h-index >= 10 • *More About the Opportunity** - Expected commitment: approximately 10–20 hours per week. - Estimated duration: ~2–3 months, with the start of 5 sample tasks in week 1 for a small pilot. • *Application Process** - Submit your CV/resume highlighting relevant research expertise. - Complete a short assessment to demonstrate ability to craft challenging, analytically complex questions. - Mercor will follow up within one week regarding next steps.
Interested in this role?Apply on iHire