AI Trainer — Verity Labs (Remote)
Trained ever-evolving large language models (LLMs) on human language patterns, benchmark testing, clarity of intention, safety metrics, and knowledge testing. Produced and reviewed model outputs using an explicit multi-page rubric to ensure consistency and quality across tasks. Worked with prompts and research inputs to assess and improve response clarity and intent alignment. • Annotated and labeled a large volume of LLM task instances against a strict rubric. • Created prompts to steer the model toward specific response goals, including humor. • Researched varied topics to evaluate knowledge coverage and the clarity of answers. • Adapted to shifting priorities with rapid turnaround across newly assigned projects.