Physics Problem Designer and AI Model Evaluator

About OpenTrain

OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. We help people discover projects, build a professional profile, and apply to opportunities in minutes. Creating an OpenTrain account is free.

OpenTrain AI is hiring a Physics Problem Designer and AI Model Evaluator for remote freelance contractor work supporting large language model training and evaluation.

• Remote work available worldwide
• Part-time contractor engagement
• Expected time commitment of 20+ hours per week
• Work in English

About AI Training and Model Evaluation

AI training is the human side of building artificial intelligence. Specialists create examples, review model outputs, write feedback, and evaluate reasoning so advanced systems can become more accurate, useful, and reliable.

In this role, your physics expertise will contribute to challenging evaluation benchmarks. Your problems, solutions, annotations, and feedback can help researchers understand where language models succeed or struggle with abstraction, symbolic manipulation, and multi-step reasoning.

• Contribute directly to the development of cutting-edge AI systems
• Use subject-matter expertise in a flexible remote setting
• Help evaluate sophisticated technical reasoning
• Build experience in a fast-growing AI training industry

The Role

As a Physics Problem Designer and AI Model Evaluator, you will create and solve challenging physics problems and assess large language model performance. Problems may span early undergraduate through PhD-level topics and should test abstraction, multi-step reasoning, and symbolic manipulation.

You will translate sophisticated physics reasoning into technically accurate, clearly structured explanations. The work also includes providing detailed annotations and constructive feedback on technical or model-generated work, while collaborating with LLM researchers to support evaluation goals and benchmark development.

• Physics-focused AI model evaluation and benchmark development
• Text-based problem design, solution writing, and technical annotation
• Advanced undergraduate through PhD-level scope
• Independent remote freelance contractor work

What You'll Do

You will design rigorous tasks that reveal the strengths and limitations of large language models. Each problem should be accompanied by an accurate, logically structured solution that makes the reasoning process clear.

You will also explain difficult concepts in accessible language, use visuals where appropriate, and provide useful feedback on technical work. Collaboration with LLM researchers will help align your contributions with evaluation goals and broad physics curricula.

• Design challenging physics problems for AI evaluation
• Solve each problem with accurate, step-by-step reasoning
• Test abstraction, multi-step reasoning, and symbolic manipulation
• Explain complex physics concepts clearly using simple language, visuals, and sound reasoning
• Review technical or model-generated work
• Provide constructive feedback and detailed annotations
• Help define benchmarks across broad physics curricula
• Collaborate with LLM researchers on evaluation objectives

Requirements

You should have strong physics foundations spanning advanced undergraduate to graduate-level topics and be able to analyze and solve complex problems using a structured approach. Clear written English is essential because the work depends on communicating technical ideas precisely.

You must be comfortable designing original problems, producing rigorous solutions, and translating sophisticated reasoning into explanations that preserve technical accuracy. Self-motivation, independence, and effective collaboration are important in this remote setting.

• Strong foundation in physics and complex problem solving
• Ability to design problems involving abstraction, multi-step reasoning, and symbolic manipulation
• Ability to produce accurate, logically structured, step-by-step solutions
• Excellent written English and clear technical communication
• Strong research, analytical, creative, and lateral-thinking skills
• Ability to give constructive feedback and detailed annotations
• Self-motivation and independence in remote work
• Desktop or laptop with a reliable internet connection

Helpful Academic Background

A current master's, PhD, or postdoctoral program in physics, applied physics, or a related field is encouraged. Candidates with experience in advanced areas of physics may be especially well suited to the breadth and difficulty of the work.

• Physics
• Applied physics
• Astrophysics and cosmology
• Nuclear physics
• Particle physics
• Related graduate-level study or research

Why Build an AI Training Career

AI training and data-labeling work offers a way to apply specialized knowledge to the systems shaping modern technology. Many projects are remote and flexible, allowing contributors to choose work that fits around other commitments while developing experience in an expanding field.

OpenTrain helps you build a lasting AI training portfolio. A strong profile can showcase credible experience, help you find projects aligned with your expertise, and support long-term growth in this industry.

• Work remotely from anywhere with a computer and internet connection
• Apply physics expertise to state-of-the-art AI development
• Develop practical experience in model evaluation and benchmark design
• Build a reusable professional portfolio through OpenTrain

How to Apply

Create your free OpenTrain account, build your profile, and apply in minutes. Highlight your physics education, research experience, problem-solving strengths, and ability to communicate advanced technical ideas clearly.

• Apply as a remote part-time contractor
• Plan for 20+ hours per week
• Complete work in English
• Use a desktop or laptop with reliable internet access