LLM Prompt-Response Evaluation for Chatbots
Worked on prompt-response evaluation for LLMs, focusing on response accuracy, coherence, tone, and safety. Followed detailed rubrics to score AI outputs and rewrote subpar responses to improve clarity and alignment with instruction. Participated in RLHF tasks including comparative ranking and fine-tuning data generation. Tasks were performed in both English and Korean, with strict guideline adherence.