Particle and Nuclear Physics AI Benchmark Task Designer

About OpenTrain

OpenTrain is the #1 platform for finding and building careers in AI training and data labeling. OpenTrain AI hires and contracts contributors for specialized projects that help shape how modern AI systems learn, reason, and perform.

• Work remotely with an international contributor community
• Build experience in a fast-growing AI training industry
• Create a profile that supports your long-term AI training career

About AI Training and Benchmark Design

AI training is the human side of building artificial intelligence. Specialists prepare examples, write challenging tasks, and evaluate model outputs so advanced systems can perform more reliably on complex real-world and research problems.

In this role, your scientific expertise will help distinguish genuine reasoning ability from surface-level pattern matching. Your work will contribute to benchmarks that test whether AI systems can carry out demanding particle and nuclear physics workflows.

• Contribute directly to the evaluation of advanced AI models
• Apply research expertise to cutting-edge computational tasks
• Work remotely with flexible scheduling

The Role

OpenTrain AI is seeking a Particle and Nuclear Physics AI Benchmark Task Designer to create and test difficult computational problems for advanced AI evaluation. You will use real particle and nuclear physics workflows, specialized scientific software, and strategic experiment design to produce graduate-level challenges that require meaningful reasoning.

Some tasks will require exact computation from a defined setup, while others will involve planning queries or experiments to uncover information that is not directly visible in the data. The role is listed at an entry level within AI training, but it requires graduate-level training or equivalent research experience in a relevant scientific field.

• Remote contract position available worldwide
• Part-time engagement of at least 15–20 hours per week
• Default weekly commitment of 40 hours
• Compensation of $70–$100 per hour
• English-language work

What You'll Do

You will design original, research-grounded computational problems and refine them through testing against advanced AI models. You will also implement the technical infrastructure needed to run and evaluate these challenges, while accounting for scientific software behavior and edge cases.

• Create graduate-level computational problems in particle and nuclear physics
• Develop tasks involving simulations, particle physics data analysis, and cross-section computations
• Design challenges involving renormalization group calculations and perturbative QCD
• Create problems related to Monte Carlo event generation and other research workflows
• Write Python problem setups, oracle functions, and solution validators
• Test problems against advanced AI models and refine them to the intended difficulty
• Separate genuine reasoning difficulty from mere computational complexity
• Identify and account for scientific software edge cases
• Design queries or experiments that reveal information not directly visible in data

Required Qualifications

This role requires strong scientific and programming expertise. You should be able to work independently on technically demanding benchmark problems, incorporate feedback, and operate in remote, terminal-based computing environments.

• Graduate-level training or equivalent research experience in particle physics, nuclear physics, or a related STEM field
• Hands-on proficiency with scikit-hep or related high-energy-physics Python libraries
• Strong Python skills for scientific computing and validation
• Experience designing or validating research-level computational problems
• Sound judgment about reasoning difficulty and software edge cases
• Familiarity with Linux or terminal-based remote compute environments
• Understanding of particle physics workflows such as cross-section calculations, perturbative QCD, or Monte Carlo event generation

Helpful Background

The following experience is helpful for producing rigorous, reproducible benchmark tasks, though the core requirements above remain central to the role.

• Benchmark or AI evaluation design
• Scientific teaching or problem-set creation
• Collider phenomenology
• Computational reproducibility
• Containerized computing environments

Work Schedule and Compensation

This is a remote contractor and part-time role. You must be available for at least 15–20 hours per week, with a default expectation of 40 hours per week. The compensation range is $70–$100 per hour.

• Employment type: Contractor
• Work arrangement: Part time
• Location: Worldwide and fully remote
• Minimum availability: 15–20 hours per week
• Default commitment: 40 hours per week
• Pay: $70–$100 USD per hour