Mathematical Reasoning & LLM Evaluation

The rapid expansion of artificial intelligence has moved far beyond simple pattern matching. Today, frontier AI research laboratories require deep logical reasoning, complex mathematical problem-solving, and rigorous scientific validation to ensure advanced models operate reliably. Bridging this gap demands elite human intellect—scholars and researchers who can translate abstract theorems and rigorous proofs into structured evaluation benchmarks.

Based in San Francisco, California, Turing stands as the world's leading research accelerator for frontier AI labs and a trusted partner for global enterprises deploying advanced AI systems. Turing empowers customers by accelerating frontier research with high-quality data, advanced training pipelines, and elite STEM researchers, while helping enterprises transform AI proofs of concept into robust proprietary intelligence.

About the Role: You Are a Fit If...

In this specialized role, you will work on cutting-edge projects that improve and evaluate large language models by applying advanced mathematical reasoning, problem-solving, computational thinking, and clear written communication. The ideal candidate possesses a robust foundation in mathematics—particularly at the level expected in rigorous engineering entrance examinations as well as graduate or PhD-level coursework.

You are an ideal fit for this engagement if you excel at breaking down intricate mathematical concepts into simple, clear explanations while working with high efficiency. This role also features computational projects where you design precise, closed-ended prompts, write reliable Python code, verify numerical answers, and provide comprehensive rationales.

What a Day-to-Day Look Like

Working as a Mathematics Expert with Turing is intellectually stimulating and dynamically varied. Your daily responsibilities include:

• Problem Generation: Designing original, highly challenging mathematics problems that test the reasoning limits of large language models in multi-step, abstract, and proof-based settings.
• Independent Solutions: Solving complex problems independently and drafting detailed, logically structured solutions with clear justifications.
• Model Evaluation: Reviewing model-generated outputs, isolating mathematical errors or missing arguments, and delivering precise annotations and corrections.
• Benchmark Development: Contributing to the creation of new evaluation benchmarks spanning early undergraduate to PhD-level mathematics curricula.
• Computational & Theorem Work: Developing and validating Python-based solutions for computational tasks and translating mathematical problems into formal language using Lean.

Perks of Freelancing With Turing

Contracting with Turing offers elite professionals unmatched career flexibility and engagement quality:

• Fully Remote Autonomy: Work from anywhere in the world with complete schedule flexibility.
• Cutting-Edge Collaboration: Direct involvement with premier LLM research labs shaping the future of artificial intelligence.
• Career Future-Proofing: Master how to leverage AI tools effectively, enhancing your analytical workflow in an AI-first economy.
• Extension Potential: Opportunities for contract extensions based on consistent performance and ongoing project needs.

Engagement Details & Onboarding Excellence

This engagement operates on flexible contractor terms (no medical or paid leave included). Candidates can select from three weekly time commitments: 20 hours/week, 30 hours/week, or 40 hours/week, requiring at least 4 hours per day with a minimum of 4 hours of overlap with Pacific Standard Time (PST).

To ensure long-term stability and unlock access to advanced, higher-tier tasks, successful applicants must complete the thorough onboarding process and complete their initial 10 hours of work with precision and dedication.