Physics Expert (PhD/Master's) for LLM Evaluation

You will create and solve physics problems that test where large language models struggle. Your work will support model evaluation and fine-tuning by measuring abstraction, multi-step reasoning, and symbolic manipulation.

The role covers physics content from early undergraduate through PhD level. You will work independently and collaborate remotely with LLM researchers to shape evaluation goals.

• Design complex physics problems that reveal weaknesses in language models.
• Solve problems in classical mechanics, electromagnetism and optics, thermodynamics, and statistical physics.
• Write accurate, clear, step-by-step solutions using structured reasoning.
• Provide detailed annotations and constructive feedback.
• Help define physics evaluation benchmarks across undergraduate and PhD-level curricula.

This is a remote, part-time contractor assignment for an individual freelancer. The role is listed as entry level, and candidates pursuing advanced physics degrees are encouraged to apply.

• 20+ hours per week.
• Remote work with a reliable desktop or laptop and good internet connection.
• Strong foundation in Physics, Applied Physics, or a related field.
• Ability to analyze and solve complex physics problems with a structured, logical approach.
• Ability to explain advanced physics concepts in simple, clear language, including through visuals and reasoning.
• Strong research, analytical, creative, and lateral-thinking skills.
• Excellent English comprehension and structured communication skills.
• Ability to work independently and collaborate effectively in a remote setting.
• A Master's, Ph.D., or postdoctoral degree in Physics, Applied Physics, or a related field is helpful but not stated as required.
• Experience with classical mechanics, electromagnetism and optics, or thermodynamics and statistical physics is especially relevant.
• The assignment may be extended based on performance and project needs.

You apply on OpenTrain with your resume and then complete the application on the hiring site.

AI training is the human work behind systems that learn from examples, including writing and rating model responses, checking reasoning, and evaluating results. People with subject expertise help make these examples accurate and useful, which is why specialized knowledge such as physics is important in this work.