AI Benchmark Scientist for Biochemistry & Genomics

Obsidian; San Francisco (CA); Remote; USD 273,000 - 382,000; Part time

Job summary

Mercor is hiring PhD and Master's scientists to author AI evaluation tasks (Sci Code) for a new benchmark in scientific computing. You will author original, executable research problems that today’s frontier models cannot solve, sourced from published papers, datasets, or open-source repositories.

The role focuses on biology-related domains with coding, including constructing prompts, defining grading criteria, and calibrating model performance.

Qualifications

• PhD in biology, biological sciences, biochemistry, genetics, ecology, or related field.
• Depth in at least two subdomains: ecology, biochemistry, genetics.
• Proficient in Python, R, or similar for scientific computing.
• Familiar with GitHub and Docker workflows.

Responsibilities

• Write scientific prompts based on sourced input (papers, Kaggle data, open-source repos).
• Source materials and design executable research problems for evaluation tasks.
• Build grading criteria to define correct answers for frontier models.
• Calibrate prompts against frontier models to ensure robust failures/successes.

Skills

Python

R

Git/GitHub

Docker

Scientific computing

Education

PhD in Biology

Tools

Docker

GitHub

Job description

Mercor is hiring PhD and Master's scientists to author AI evaluation tasks (Sci Code) for a new benchmark in scientific computing. You will author original, executable research problems that today’s frontier models cannot solve, sourced from published papers, datasets, or open-source repositories.

The role focuses on biology-related domains with coding, including constructing prompts, defining grading criteria, and calibrating model performance.