Applied Physicist — AI Physics Auditor (Remote)

Alignerr Corp.; Boston (MA); Remote; USD 83,000 - 165,000; Full time

Benefits offered by this job

Fully remote work

Flexible hours

Job summary

Alignerr Corp. is seeking PhD-level Applied Physicists to stress-test cutting-edge Large Language Models, exposing gaps in physical reasoning and ensuring models respect first principles.

This fully remote hourly contract invites researchers to design advanced problems, author golden solutions with exact constants and units, audit AI reasoning, and refine model behaviour across classical mechanics, electromagnetism, statistical mechanics, and quantum mechanics.

Qualifications

• PhD in Applied Physics, Physics, Engineering Physics or closely related field.
• Exceptional analytical writing and ability to communicate complex derivations clearly.
• Self-directed and reliable working independently on technical tasks.
• No prior AI or data annotation experience required.

Responsibilities

• Design Advanced Problems—craft complex physics problems at PhD qualifying exam level.
• Author Ground-Truth Solutions—provide rigorous, step-by-step golden responses with correct constants and units.
• Audit AI Reasoning—evaluate AI-driven simulations and proofs for physical consistency.
• Refine Model Behaviour—offer feedback to improve reasoning within conservation laws and symmetry principles.
• Cover Core Domains—work across Classical Mechanics, Electrodynamics, Statistical Mechanics, and Quantum Mechanics.

Skills

Analytical writing

Self-directed work

Rigorous with units

English communication

Education

PhD in Applied Physics

Tools

MATLAB

COMSOL

Python (NumPy/SciPy)

Job description

Alignerr Corp. is seeking PhD-level Applied Physicists to stress-test cutting-edge Large Language Models, exposing gaps in physical reasoning and ensuring models respect first principles.

This fully remote hourly contract invites researchers to design advanced problems, author golden solutions with exact constants and units, audit AI reasoning, and refine model behaviour across classical mechanics, electromagnetism, statistical mechanics, and quantum mechanics.