Applied Physics Benchmark Specialist

$61–77/hrrate as listed

Write and review expert-level physics multiple-choice questions — one correct answer, nine subtly wrong distractors, and a full chain-of-thought solution — for use as AI benchmark items. Work is asynchronous and remote at roughly 10+ hours per week, with observed pay of $61–77/hr.

Summary, eligibility guidance and screening preparation are prepared by Sidequest from the platform's posting. Check the source listing before applying.

What the work involves

You will be assigned to one of two tracks. Question authoring means writing original problems in your subdomain — semiconductor physics, nanoelectronics and spintronics, photonics, quantum optics and ultrafast, quantum sensing and metrology, plasma and fusion, nonlinear dynamics and turbulence, or geophysics and reservoir simulation. Each item needs a self-contained statement with every constant, boundary condition, and convention specified; one defensible correct answer; nine distractors that a strong physicist could plausibly land on after a specific mistake; a markdown chain-of-thought solution with clean intermediate steps; a difficulty rating (Medium / Hard / Expert); and one to five references from peer-reviewed journals, textbooks, or university repositories.

Question verification means working the other direction: solving pre-written items independently, then auditing them for ambiguity, missing information, unsolvability, sign or unit errors, and distractors that are either obviously wrong or accidentally also correct. You edit, re-rate difficulty, and document what you changed and why. Reviewers are generally held to the standard that a rejection or edit must be justified in physics terms, not stylistic preference.

What the platform screens for

Mercor's screening is AI-led and leans on verifiable specifics. Expect your CV and publication record to be probed for the subdomain you claim — which formalism you actually use, what you measured or simulated, what your thesis argued. Then expect judgment probes: given a flawed item, can you say precisely what is wrong with it, and can you construct a distractor that encodes a named misconception rather than a random perturbation? Writing quality matters as much as physics depth here; the deliverable is prose and LaTeX-adjacent markdown, and unclear phrasing is the most common rejection reason on benchmark work.

Logistics

Fully remote, fully asynchronous — no standing meetings, work drawn from a queue
Expected commitment around 10+ hours/week; pay observed in the $61–77/hr band, typically varying with subdomain scarcity and track, and not guaranteed
PhD or doctoral candidate in physics, applied physics, astrophysics, or adjacent; a master's can clear the bar with exceptional depth in one named subdomain
Prior olympiad problem writing, qualifying-exam authoring, or textbook problem-set work is a meaningful advantage

What they look for

PhD or doctoral candidacy in physics, applied physics, astrophysics, or closely related field
Deep specialization in at least one listed subdomain (e.g. spintronics, quantum metrology, plasma/fusion, turbulence, reservoir simulation)
Graduate-level mathematical formalism and the ability to carry a derivation cleanly end-to-end
Rigorous problem design — unambiguous statements, calibrated difficulty, misconception-based distractors
Clear technical writing in English, including markdown-formatted step-by-step solutions
Sourcing and citing peer-reviewed literature or reputable academic repositories

Before you apply

Would you pass this screen?

Drawn from what this listing actually requires. Answered in your browser — nothing is sent to us, and we do not store your answers.

Do you hold a PhD in Physics, Applied Physics, Astrophysics, or a closely related field, or are you a current doctoral candidate?
PhD or equivalent in Physics or related field · required
Can you solve multi-step problems involving graduate-level concepts (e.g., quantum mechanics, electromagnetism, thermodynamics) and explain your reasoning in rigorous mathematical terms?
Graduate-level physics knowledge · required
Have you authored, reviewed, or curated physics exam questions, problem sets, competition problems, or similar rigorous academic assessments?
Academic problem-design or assessment experience · preferred
Are you able to write technical physics content in clear, concise English without a need for significant editing for clarity or grammar?
Strong written English · preferred
Can you commit to working at least 10 hours per week on this asynchronous project, with the expectation that your involvement will continue for multiple months?
10+ hours per week availability for several months · required

0 of 5 answered

Included with this listing

Interview prep — 11 questions

Matched to what this platform's screening actually measures. Answer each out loud in 60–90 seconds: claim → concrete example → implication.

What it tests: Can you design plausible distractors that exploit real misconceptions, not arbitrary wrong answers? This separates benchmark-quality from mediocre assessment.
A strong answer: A strong answer names a specific question or domain, describes the conceptual error the distractors target (e.g., sign error from boundary conditions, or dimensional analysis trap), and shows how each wrong answer maps to a genuine mistake a skilled solver might make.
Where people lose it: Saying 'I made sure the wrong answers were all clearly wrong' — distractors must be believable to someone who knows the domain but made a reasoning slip. Generic 'I picked plausible numbers' shows no grasp of the real craft.