Principal Research Scientist (Computer Science / Computational Physics)

$100–400/hr

rate as listed

Workload: ~15h/week
Engagement: Project-based
Location: Worldwide
Field: Sciences

Evaluate frontier AI research in computer science or mathematical and computational physics through proofs, algorithms, numerical experiments, and simulation. The work centers on defining tractable open problems, validating model outputs, and documenting genuine progress and failure modes.

The work

You will frame foundational open problems, theoretical bounds, or simulation challenges for AI models and agents, then guide multi-step research involving proofs, algorithms, numerical experiments, or high-performance simulations. Day-to-day outputs may include reproduced computations, proof checks, novelty assessments, and concise research memos covering partial solutions, failed approaches, and model limitations.

What the platform screens for

The screen is likely to focus on verifiable research contributions, command of current literature, and the ability to distinguish plausible output from correct and significant work. Strong candidates can explain a named result in detail, describe how they would reproduce or falsify an AI-generated claim, and identify research questions with measurable success criteria. Experience with Lean, Coq, SAT/SMT solvers, program committees, editorial boards, or public AI-assisted research is a plus, not a stated requirement. No prior AI experience is required.

Logistics

This is a remote contractor role with flexible part-time or full-time participation. The listing does not specify fixed weekly hours, time-zone coverage, or synchronous meeting requirements, so candidates should be ready to state realistic availability and any computational resource constraints. The platform's published band is observational, not guaranteed.

Pay band: $100–400/hr

What they look for

Novelty judgment
Research direction
Computational rigor
Complex problem solving - Research

Included with this listing

Interview prep — 10 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: This tests whether you can connect a specific, recent scholarly result to a precise technical contribution rather than relying on titles, affiliations, or publication counts.
A strong answer: Name the result, venue, date, your individual contribution, the prior limitation it addressed, and the evidence supporting the claim. Be prepared to explain the central method and what remains unresolved.
Where people lose it: Candidates lose points by describing a broad research program without separating their own contribution from that of collaborators.

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