- Experience
- Any
- Salary
- —
- Openings
- 1
- Posted
- 3 weeks ago
- Work mode
- In office
- Education
- Ph.D.
- Resume
- Required to apply
Sign in to tell us what does and doesn't work for you here — it sharpens every match we show you.
Job description
Role Overview
Seeking a Mathematics expert to craft realistic, terminal-based scientific computational tasks focused on Terminal Bench Science. Responsibilities include designing multi-step problems in numerical analysis, optimization, statistics, mathematical modeling, probability, dynamical systems, and computational mathematics. The goal is to assess if AI agents can formulate mathematical challenges, implement dependable algorithms, execute them in terminal environments, debug numerical flows, and deliver verifiable outputs.
Core Domains
Applied mathematics involving numerical analysis, optimization, statistics, probability, mathematical modeling, differential equations, dynamical systems, operations research, and computational geometry.
Responsibilities
- Design authentic, complex computational mathematics problems.
- Convert mathematical workflows and research processes into standalone terminal environments.
- Prepare related datasets, equations, model parameters, constraints, initial conditions, and expected outcomes.
- Develop expert-level solutions utilizing Python, R, Julia, C/C++, Bash or other technologies.
- Create tasks on optimization, numerical integration, differential equations, matrix computations, statistical inference, stochastic modeling, and algorithm analysis.
- Define strict grading standards emphasizing numerical accuracy, convergence, computational complexity, feasibility, and mathematical precision.
- Set suitable tolerances, stopping criteria, stability checks, and reproducibility controls.
- Develop automated validation tests covering edge cases and alternative valid results.
- Troubleshoot floating-point precision challenges, solver issues, convergence barriers, conditioning, and performance concerns.
- Document underlying assumptions, mathematical structures, expected results, and any known task limitations.
Requirements
- Ph.D., postdoctoral experience, or equivalent advanced qualifications in mathematics, statistics, or related fields.
- Strong programming capabilities in Python, R, Julia, C/C++, Bash, or similar languages.
- Experience using Linux or terminal-based development environments.
- Proficient with numerical methods, optimization, statistical analysis, mathematical modeling, and scientific computations.
- Ability to independently code, test, and verify computational algorithms.
- Deep knowledge of numerical stability, error estimation, mathematical assumptions, and reproducibility procedures.
Preferred Qualifications
- Experience with libraries such as NumPy, SciPy, SymPy, pandas, scikit-learn, JAX, PyTorch, CVXPY, or related tools.
- Familiarity with optimization solvers, probabilistic programming frameworks, differential-equation solvers, or numerical linear algebra utilities.
- Hands-on with Monte Carlo simulations, stochastic modeling, Bayesian inference, graph algorithms, or computational geometry.
- Knowledge of Docker, Conda environments, Git version control, continuous integration systems, and automated testing pipelines.
- Background in developing mathematical programming challenges, benchmark problems, or automated grading systems.
Additional Preferences
- Experience assessing AI-driven coding or terminal-based agents.
- Research software engineering background.
- Expertise in high-performance computing or optimization of performance-critical numerical code.
- Contributions to published applied mathematics or computational science research or open-source projects.
- Skill in designing tests distinguishing mathematically valid solutions from approximate or unstable numeric solutions.
Minimum education
Doctorate