Work Type
Remote
Location
Global
Compensation
Up to $70/hr
About this opportunity
Overview
Mercor is seeking physicists to join an AI training project. In this contract role, you will review AI-generated answers, assess the quality of reasoning and step-by-step problem solving, and provide expert feedback to help improve model accuracy, logic, and clarity. The interview language is English.
What You'll Do
- Review AI responses for correctness, clarity, and alignment with the prompt.
- Identify mistakes in calculations, methodology, and conceptual understanding.
- Fact-check physics content using public sources.
- Write high-quality physics explanations and model solutions that use correct methods.
- Rate and compare multiple AI outputs based on physics accuracy and reasoning quality.
- Create detailed prompts and responses to support AI training across a range of topics.
- Evaluate and rank AI responses to improve fluency, accuracy, and contextual relevance.
- Test models for inaccuracies or bias and help validate reliability across use cases.
Requirements
- MS or PhD in Physics or a related field from a top 100 university.
- Strong grounding in mechanics, electromagnetism, optics, thermodynamics/statistical mechanics, and modeling/systems thinking.
- Experience developing or critically reviewing complex physics content such as problem banks, proofs, textbook sections, or research notes.
- Ability to fact-check information using public sources across multiple domains.
- Clear physics communication and strong scientific writing ability.
- Background in research, analytical writing, debate, programming, or physics is preferred.
- Published in a peer-reviewed journal.
- C1-level English proficiency or above.
- Bonus experience in data labeling, RLHF, or AI model evaluation.
Skills
- Physics reasoning
- Computational physics
- Experimental physics
- Theoretical physics
- Research
- Scientific writing
- AI model evaluation