Pareto.AI is a human data collection platform connecting leading AI researchers with trusted industry experts to collaborate on AI alignment, safety, and training projects. By working together, we can better align AI models with human values and develop more helpful, honest, and harmless AI models.

We are currently seeking Mathematics Experts to contribute their expertise enhancing LLM performance for an AI training project. This exciting opportunity allows you to play a key role in advancing the accuracy and reliability of cutting-edge AI systems. Someone who has competed in a Mathematics Olympiad would be ideal!

PROJECT OVERVIEW: Enhancing LLM Performance in Mathematics

This project focuses on improving the performance of large language models (LLMs) in Mathematics by ensuring they are trained on high-quality math datasets. To achieve this, we are assembling a team of experts with strong mathematical backgrounds to do the following:

Scope of Work:

✅ Cross-checking data extracted by models to ensure accuracy.

✅ Verifying reference answers to confirm correctness and alignment with the given question.

✅ Assessing difficulty levels (e.g., Competition, University, High School).

✅ Solving problems independently—experts must solve all problems themselves and are strictly prohibited from using ChatGPT or any other AI tools for assistance.

This fully remote project is open to participants based in the US, Canada, UK, Australia, and New Zealand. Compensation is approximately $40 USD per approved hour. The work schedule is flexible, with an expected commitment of 20-40 hours per week for the duration of the project.

For immediate consideration, please apply now!