Physics and Mathematics Expert Contributor, Outlier AI
Worked as a Physics and Mathematics Expert Contributor on projects evaluating and improving LLM-generated responses to complex prompts in physics and mathematics. Responsibilities included engineering open-ended, domain-specific prompts designed to expose model failures, and creating rubrics to assess AI performance. Actively participated in iterative model evaluation and improvement cycles with expert peers. • Evaluated model responses to complex subject matter prompts. • Engineered prompts to intentionally induce subpar model output. • Contributed to rubric creation for objective evaluation. • Worked on continuous improvement of LLM models in STEM domains.