Remote AI Evaluation Scientist — Advanced STEM Problems

24-Mag Llc; New York (NY); Remote; USD 80,000 - 150,000; Full time

Job summary

24-MAG LLC invites experts in advanced STEM fields to design and solve challenging problems for large language model evaluation. This remote, contractor-style engagement emphasizes rigorous reasoning, structured solutions, and detailed annotations, spanning physics, mathematics, biology, chemistry, and related disciplines.

Ideal candidates hold a Master’s, Ph.D., or postdoctoral position and demonstrate strong analytical and research capabilities, with clear English communication and the ability

Qualifications

• Must be pursuing or holding a Master’s, Ph.D., or postdoctoral position in a STEM field.
• Strong analytical and research capabilities with ability to solve complex quantitative and conceptual problems.
• Excellent written English communication and ability to explain technical concepts clearly.

Responsibilities

• Design challenging STEM problems across advanced domains for evaluation.
• Produce clear step-by-step solutions with explicit assumptions.
• Evaluate model responses and identify reasoning weaknesses.
• Contribute to evaluation design and benchmarking across disciplines.
• Collaborate remotely with AI researchers and provide constructive feedback.

Skills

Analytical thinking

Research skills

STEM reasoning

Explanations clarity

Problem solving

English communication

Remote collaboration

Self-motivation

Education

Master's/PhD/postdoc in STEM

Tools

Python

LaTeX

Job description

24-MAG LLC invites experts in advanced STEM fields to design and solve challenging problems for large language model evaluation. This remote, contractor-style engagement emphasizes rigorous reasoning, structured solutions, and detailed annotations, spanning physics, mathematics, biology, chemistry, and related disciplines.

Ideal candidates hold a Master’s, Ph.D., or postdoctoral position and demonstrate strong analytical and research capabilities, with clear English communication and the ability