The Work
You will help evaluate large language models using advanced mathematics problems, model-generated solutions, computational tasks, and formal proofs. Your work will help create structured examples and clear feedback for AI systems.
- Design challenging mathematics problems covering multi-step, abstract, and proof-based reasoning.
- Solve problems independently and write logically structured solutions with clear justifications.
- Review model-generated solutions for mathematical errors, missing arguments, and weak reasoning.
- Provide precise feedback, annotations, and corrections.
- Contribute to evaluation benchmarks covering early undergraduate through doctoral-level mathematics.
- Develop and validate Python solutions with approved scientific libraries.
- Translate mathematical problems and proofs into Lean and verify that formal proofs compile correctly.
What It Pays and Takes
This is a fully remote, part-time contractor assignment. The listing is marked entry level, while the work requires strong mathematics knowledge and careful written analysis.
- Pay: No rate is provided in the listing.
- Commitment: At least 20 hours per week.
- Available schedules: 20, 30, or 40 hours per week.
- Daily schedule: At least 4 hours per day, including 4 hours of overlap with Pacific Time.
- Location: Open worldwide.
- Language: Fluent in English.
- Education or experience: Relevant academic or professional expertise in mathematics, applied mathematics, statistics, or a related field.
- Knowledge: Strong foundations in areas such as algebra, calculus, analysis, geometry, or topology.
- Communication: Ability to explain complex mathematical ideas clearly with structured language and examples.
- Technical skills: Python and scientific libraries are valuable; familiarity with Lean or formal proof verification is beneficial.
- Work style: Strong written communication, attention to detail, independent work habits, and the ability to follow detailed project guidelines.
How It Works
Apply on OpenTrain with your resume, then complete the application on the hiring site.
About AI Training Work
AI training is the human work behind systems that learn from examples, including reviewing model answers, solving problems, and rating the quality of responses. People with specialized knowledge are paid for careful judgment that helps make AI systems more accurate and useful.