Math Expert or Specialist

Mathematical evaluation

Design rigorous problems that help train and assess large language models on advanced quantitative reasoning. Assignments span undergraduate mathematics through doctoral-level material.

• Develop problems requiring several reasoning steps, including abstraction, symbolic manipulation, and proof construction.
• Write reference solutions with a clear logical structure and fully explained steps.
• Evaluate generated work for mathematical correctness, reasoning depth, and solution quality.
• Work with researchers to broaden and refine benchmark collections.
• Provide detailed feedback showing how models can improve on complex quantitative tasks.

Required expertise

You must be pursuing or hold a PhD in Mathematics, Applied Mathematics, Statistics, or a closely related discipline. Strong reasoning across advanced subjects, including analysis, algebra, and probability, is essential.

You should recognize incomplete or invalid logic and explain corrections precisely in writing. Independent, asynchronous assignments call for a careful approach and dependable delivery; AI or machine learning experience is not required.

Helpful background

Annotation, quality assessment, and evaluation systems experience is useful. Familiarity with AI tools, breadth across pure, applied, computational, or statistical mathematics, and experience in academic research, tutoring, or curriculum design are welcome.

Contract scope

This fully remote hourly contract involves 10 to 40 hours per week with flexible scheduling. It offers direct exposure to how language models are trained and evaluated alongside AI researchers, with possible continuing assignments or contract extension.