AI Trainer & AI Model Evaluator
As an AI Trainer & AI Model Evaluator, I evaluated and improved large language model (LLM) responses for accuracy and instruction adherence. My responsibilities included prompt-response evaluation, reinforcement learning with human feedback (RLHF), and developing annotation guidelines. I ensured high-quality outputs by conducting hallucination detection and bias analysis. • Evaluated LLM prompts and responses for accuracy and factuality. • Ranked AI-generated outputs based on helpfulness. • Created and refined evaluation rubrics for annotation tasks. • Flagged edge cases and facilitated model safety improvements.