Physics Expert for LLM Evaluation

About OpenTrain

OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. OpenTrain AI recruits and contracts contributors for specialized projects where human expertise helps improve the next generation of artificial intelligence.

This assignment is a remote freelance opportunity for an individual contractor. You can create an OpenTrain account for free, build your AI training profile, and apply in minutes.

• Remote work available worldwide
• Contractor and part-time arrangement
• 20+ hours per week
• Work conducted in English

About AI Training and LLM Evaluation

Large language models learn from examples, feedback, and evaluations prepared by people. In this role, your physics expertise will help test whether models can handle abstraction, multi-step reasoning, symbolic manipulation, and advanced technical concepts.

AI training work is a growing area of tech that can offer flexible, remote opportunities. Contributors directly shape how AI systems solve problems, explain ideas, and respond to difficult questions.

• Contribute to language-model evaluation and fine-tuning
• Apply subject-matter expertise to cutting-edge AI development
• Work remotely with flexible freelance scheduling

The Role

OpenTrain is seeking a Physics Expert for LLM Evaluation to design and solve challenging physics problems that expose the limitations of large language models. You will create rigorous, step-by-step solutions, provide detailed reasoning and feedback, and contribute to evaluation benchmarks spanning early undergraduate through PhD-level physics.

The work covers physics evaluation objectives across classical mechanics, electromagnetism and optics, and thermodynamics and statistical physics. You will work independently while collaborating remotely with LLM researchers to align problem sets with benchmark goals.

• Create challenging physics problems for language-model evaluation
• Support benchmarks spanning undergraduate through PhD-level curricula
• Focus on multi-step reasoning, abstraction, and symbolic manipulation

What You'll Do

You will develop technically demanding content and explain your reasoning clearly enough for evaluation and annotation workflows. Your feedback will help identify model weaknesses and improve the quality of training and evaluation data.

• Design complex physics problems that reveal where language models struggle
• Solve problems in classical mechanics, electromagnetism and optics, and thermodynamics and statistical physics
• Write clear, accurate, step-by-step solutions using structured reasoning
• Collaborate with LLM researchers to align problem sets with evaluation objectives
• Provide constructive feedback and detailed annotations
• Help define physics evaluation benchmarks across multiple academic levels

Requirements

You should have a strong or graduate-level foundation in Physics, Applied Physics, or a related field. The work requires the ability to solve complex problems rigorously and communicate advanced concepts in simple, clear language.

Candidates currently pursuing a Master's, Ph.D., or postdoctoral degree in Physics, Applied Physics, or a related field are encouraged to apply. Experience in classical mechanics, electromagnetism and optics, or thermodynamics and statistical physics is especially relevant.

• Strong foundation in Physics, Applied Physics, or a related field
• Ability to analyze and solve complex physics problems with structured logic
• Ability to produce rigorous solutions involving multi-step reasoning and symbolic manipulation
• Ability to explain advanced physics concepts clearly using simple language, visuals, and reasoning
• Strong research, analytical, creative, and lateral-thinking skills
• Excellent English comprehension and structured communication skills
• Ability to provide constructive feedback and detailed annotations for language-model evaluation
• Reliable desktop or laptop with a good internet connection

Freelance Arrangement

This is a remote contractor assignment for an individual freelancer. The work may be extended based on performance and project needs, and the assignment is structured for part-time participation of 20 or more hours per week.

• Worldwide eligibility
• Individual freelance contractor arrangement
• Part-time commitment of 20+ hours per week
• Potential extension based on performance and project needs

How to Apply Through OpenTrain

Create a free OpenTrain account, build your profile around your physics background, and apply in minutes. Be ready to demonstrate your ability to reason through advanced problems, produce accurate step-by-step solutions, and communicate technical ideas clearly.

• Create your free OpenTrain account
• Highlight your physics education and relevant subject expertise
• Apply for this remote AI training assignment
• Use your profile to continue building an AI training career