Physical Sciences Expert

We are looking for Physical Sciences Experts to develop realistic, terminal-based scientific tasks for Terminal Bench Science. The role involves translating authentic workflows from physics, chemistry, materials science, astronomy, and computational science into reproducible computational environments.

You will contribute to the development of scientific inputs, computational models, executable solutions, automated tests, and objective grading criteria to evaluate AI agents on scientific reasoning, command-line operations, debugging, and the creation of reliable scientific outputs.

Key Responsibilities

• Design multi-step terminal-based tasks based on realistic physical-science workflows.
• Develop self-contained computational environments with scientific software and pinned dependencies.
• Create datasets, molecular structures, simulation parameters, experimental data, and model configurations.
• Implement expert solutions using Python, Bash, C/C++, Julia, or relevant domain-specific tools.
• Develop automated tests for numerical accuracy, physical consistency, convergence, and output structure.
• Create tasks involving simulations, numerical modeling, data fitting, optimization, spectroscopy, molecular analysis, and scientific visualization.
• Define appropriate numerical tolerances, units, boundary conditions, and expected scientific behavior.
• Validate that tasks are reproducible and execute successfully without runtime downloads.
• Debug issues related to dependencies, numerical precision, solver stability, performance, and file formats.
• Clearly communicate scientific assumptions and computational limitations to reviewers.

Required Qualifications

• Ph.D., postdoctoral experience, or equivalent advanced technical experience in a relevant physical-science discipline.
• Strong programming skills in Python, C/C++, Julia, Bash, or another scientific programming language.
• Experience working in Linux or terminal-based environments.
• Experience with numerical methods, scientific modeling, simulations, or quantitative data analysis.
• Ability to independently create and validate computational scientific workflows.
• Strong understanding of units, numerical precision, physical constraints, and scientific reproducibility.

Relevant Domains
Candidates with expertise in one or more of the following areas are encouraged to apply:

• Physics
• Chemistry
• Materials Science
• Computational Physics
• Computational Chemistry
• Astronomy & Cosmology
• Thermodynamics
• Quantum Mechanics
• Statistical Mechanics
• Materials Modeling

Preferred Qualifications

• Experience with NumPy, SciPy, pandas, matplotlib, SymPy, JAX, or similar libraries.
• Familiarity with molecular dynamics, quantum chemistry, finite-difference methods, Monte Carlo methods, optimization, or statistical modeling.
• Experience with scientific tools such as OpenMM, ASE, RDKit, Psi4, LAMMPS, GROMACS, or similar domain-specific software.
• Familiarity with Docker, Conda, Git, CI systems, and automated testing.
• Experience working with HPC systems or performance-sensitive scientific workloads.

Additional Advantage

• Experience evaluating AI coding or terminal agents.
• Research Software Engineering experience.
• Experience creating benchmark tasks or automated graders.
• Publications or open-source contributions involving computational science.
• Experience converting experimental or research workflows into reproducible packages.

Engagement Details

• Employment Type: Contractor Assignment
• Duration: 5 Weeks
• Commitment: 40 hours per week
• Time Zone Requirement: 4 hours of overlap with PST per day
• Payment: $300 per approved task
• Start: Immediate

Selection Process
Candidates will complete:

• Terminal Bench Screening Test 20 minutes (Mandatory)
• One 40-minute Physical Sciences assessment in one of the following areas:
• Astronomy & Cosmology
• Physics
• Materials Science
• Chemistry

Total assessment time: 60 minutes.