PhD Physicist – Frontier Reasoning Trajectory Writer (AI Trainer)

About the role

Mercor is looking for a PhD‑qualified physicist to create high‑quality "golden trajectories" – step‑by‑step solution walkthroughs for challenging physics problems. You will work remotely, collaborating with a frontier large language model (LLM) to draft, refine, and polish reasoning paths that can be used as training data.

Key responsibilities

• Produce complete, publication‑grade derivations for hard physics problems using LaTeX.
• Run iterative LLM workflows, providing hints, spotting reasoning errors, and shaping clean trajectories.
• Perform numerical sanity checks and generate plots with Python.
• Deliver roughly one to two tasks per week, each averaging about 7 hours of work.

Required profile

• PhD in physics or within the final 1‑2 years of a PhD program from a reputable institution.
• Specialisation in theoretical or computational astrophysics, condensed‑matter theory, quantum physics/information, or particle/high‑energy physics.
• Strong attention to detail and ability to produce publication‑standard derivations.

Required skills

• Fluency in LaTeX for mathematical derivations.
• Proficiency in Python for numerical checks and plotting.
• Experience with iterative LLM prompting and error correction.

What we offer

• Fully remote, flexible schedule (≈15 hours/week).
• Independent‑contractor arrangement with weekly payments via Stripe or Wise.
• Opportunity to work on cutting‑edge AI‑training data for physics.