Remote Physics Evaluator & Benchmark Designer

Crossing Hurdles; United States; Remote; USD 55,000 - 83,000; Full time

Job summary

Crossing Hurdles is seeking researchers to create and solve advanced physics problems designed to evaluate and improve large language models. You will develop clear, step-by-step solutions and high-quality benchmark data.

You will analyze AI-generated responses, collaborate with AI/LLM researchers, cover topics from undergraduate to PhD level, and provide detailed feedback. This remote role requires a personal laptop, reliable internet, and availability for at least 4 hours per day with PST

Qualifications

• PhD or Doctorate in physics or equivalent field.
• Strong knowledge of physics and advanced problem-solving.
• Excellent analytical and research skills with structured reasoning.
• Strong written English and ability to explain complex physics concepts clearly.

Responsibilities

• Create and solve advanced physics problems designed to evaluate and improve large language models.
• Develop clear, accurate, step-by-step solutions and high-quality benchmark data.
• Analyze AI-generated responses and identify errors in reasoning, calculations, and concepts.
• Collaborate with AI/LLM researchers to design evaluation tasks and benchmarks.
• Work across physics topics ranging from advanced undergraduate to PhD-level concepts.
• Provide detailed feedback and annotations to support AI model improvement.
• Work independently in a remote, task-based environment while meeting quality and delivery requirements.

Skills

Strong physics knowledge

Advanced problem solving

Analytical reasoning

Excellent written English

Experience in AI evaluation or data

Familiar with AI/LLM tools

Clear scientific communication

Independent remote work

Education

PhD in Physics, Applied Physics, or equivalent

Job description

Crossing Hurdles is seeking researchers to create and solve advanced physics problems designed to evaluate and improve large language models. You will develop clear, step-by-step solutions and high-quality benchmark data.

You will analyze AI-generated responses, collaborate with AI/LLM researchers, cover topics from undergraduate to PhD level, and provide detailed feedback. This remote role requires a personal laptop, reliable internet, and availability for at least 4 hours per day with PST