AI Content Evaluator
As an AI Content Evaluator, I conducted reinforcement learning from human feedback to optimize large language models. My work involved adversarial testing to identify model hallucinations and enhance safety mechanisms. I developed and refined prompts to assess model reasoning in various domains, ensuring data quality and guideline adherence. • Evaluated LLM outputs for factual accuracy and logical consistency. • Performed model auditing for safety compliance using RLHF. • Designed and implemented test prompts for creative, technical, and mathematical reasoning. • Utilized Data Annotation Tech's proprietary platform for all labeling tasks.