PhD Physicist for AI Model Training [Remote]

In this role, you will create "golden trajectories" for challenging physics problems by providing complete, step-by-step solution walkthroughs. Each trajectory will ensure that every reasoning step is sound, the final answer is accurate, and the work is organized around a coherent upfront plan. You will collaborate in a studio environment, iterating with a frontier language model to draft and refine your reasoning, ultimately producing trajectories that are polished enough to serve as training data. The problems you will tackle are rigorously selected, requiring the frontier language model to fail at least once during screening. Domain Focus

We are looking for physicists with a theoretical or computational specialization in one of the following areas:

• Astrophysics (theory / numerical)
• Condensed matter (theory)
• Quantum physics / quantum information
• Particle / high-energy physics

We are not considering experimentalists, software engineers, or data scientists without a physics PhD background.

• PhD in physics, or in the final 1-2 years of a PhD program, from a credible research institution
• Daily-driver fluency in LaTeX for derivations and Python for numerical sanity checks and plotting
• Comfort with iterative language model workflows, including nudging models with hints, identifying reasoning errors, and shaping explorations into clean trajectories
• Detail-oriented with the ability to produce publication-grade derivations

Approximately 15 hours per week, with an expectation of one to two tasks per week. Each task typically requires about 7 hours of writing time.

Hourly compensation ranges from $80 to $120.

This position is open to individuals with the required qualifications and experience.