Remote Biology AI Evaluation Scientist

24-Mag Llc; New York (NY); Remote; USD 70,000 - 100,000; Full time

Benefits offered by this job

Fully remote work

Opportunity for contract extension

Flexible schedule

Job summary

24-MAG LLC is offering a fully remote, contractor-style engagement for advanced biology researchers to design complex problems and develop structured, multi-step solutions for AI model evaluation. You will shape benchmarks across undergraduate to PhD-level biology and collaborate with research teams on evaluation strategies.

The role emphasizes rigorous reasoning, data interpretation, and clear annotations, with opportunities for project extensions depending on performance and requirements.

Qualifications

• Pursuing or holding a Master's/Ph.D./postdoc in Biology or related field.
• Backgrounds may include Biotechnology, Biochemistry, neuroscience, or related areas.
• Strong analytical and research capabilities; explain complex concepts.
• Experience with data interpretation and structured problem solving.

Responsibilities

• Design challenging biology problems across advanced domains.
• Solve complex biology problems with clear, rigorous step-by-step solutions.
• Evaluate language-model performance on biology tasks and identify weaknesses.
• Contribute to new biology evaluation benchmarks and evaluation criteria.
• Collaborate remotely with researchers and communicate findings clearly.

Skills

Biology research

Problem design

Data interpretation

Technical writing

Remote collaboration

Education

Master's/Ph.D./postdoc in Biology

Job description

24-MAG LLC is offering a fully remote, contractor-style engagement for advanced biology researchers to design complex problems and develop structured, multi-step solutions for AI model evaluation. You will shape benchmarks across undergraduate to PhD-level biology and collaborate with research teams on evaluation strategies.

The role emphasizes rigorous reasoning, data interpretation, and clear annotations, with opportunities for project extensions depending on performance and requirements.