AI Data Trainer - Text and RLHF Annotation
As an AI Data Trainer, I consistently test and refine prompts to evaluate AI model comprehension and reasoning. My work involves assessing and rating AI responses for accuracy, fluency, tone, bias, and consistency. I generate, label, and rate diverse training data such as Q&A pairs, instructions, and code samples to improve AI models. • Evaluated outputs and provided structured feedback for model tuning and RLHF. • Created original text datasets for question answering and instruction following. • Ranked model completions for logical coherence and absence of hallucinations. • Tested large language models across multiple AI platforms including ChatGPT and Claude.