AI Training Evaluator (Freelance Contract)
As an AI Training Evaluator at Outlier, I reviewed, labeled, and refined outputs from large language models to improve accuracy. My work involved debugging code, categorizing AI responses, and developing robust training data and rubrics. Rigorous assessment ensured feedback alignment and enhanced future model tuning. • Provided detailed evaluations and corrections for AI-generated content. • Debugged and categorized multiple code segments. • Developed and refined data quality rubrics and realistic sample data. • Ensured annotation protocols were strictly followed for training optimization.