Physics PhD (Top 50 Program, US/UK/CA)
Obsidian
Compensation
$70-$180/hr
Description
Role Overview
Mercor is collaborating with a leading AI research group to engage experts with advanced training in physics. PhD graduates in Physics from a US/UK/CA top 50 pure physics program are encouraged to participate. This cutting-edge project focuses on strengthening AI systems’ reasoning capabilities through rigorous problem-solving and theorem-based evaluation. You will be applying graduate-level expertise to assess and validate complex mathematical reasoning.
Key Responsibilities
-
Solve and evaluate challenging physics problems across a range of subfields
-
Review AI-generated proofs, derivations, and solutions for correctness, rigor, and clarity
-
Identify gaps, inconsistencies, or unsupported reasoning in model outputs
-
Apply relevant citations, principles, and formal reasoning where appropriate
Ideal Qualifications
-
PhD in Physics from a top 50 physics program in the United States, United Kingdom, or Canada
-
Degree must be in Physics rather than applied physics, engineering, or a closely related applied discipline
-
Strong graduate-level problem-solving ability across advanced physics topics
-
Ability to clearly articulate reasoning, reference relevant theorems or principles, and identify gaps in written solutions
-
Ideally active in research or academia
-
Able to commit approximately 15 hours per week
-
Able to complete assigned tasks in a single sitting within 24 hours when work is available
-
Available to begin immediately
More About the Opportunity
-
Work is project-based and delivered in batches
-
There may be periods when no work is available
-
Contributors should be comfortable with variable task volume depending on project needs
Application Process
-
Submit your resume and complete an interview to get started (if we do not have one already on file)
-
Qualified applicants will receive follow-up within a few days with next steps
- Commitment
- Hourly
Skills & categories
- Posted
- Jun 29, 2026
- Slots remaining
- 51
- First seen
- Jun 30, 2026
- Last seen
- Jun 30, 2026