AI Training & Data Annotation coursework (LLM evaluation, RLHF, and responsible data use)
Text
RLHFCompleted structured learning focused on LLM alignment and data-centric training workflows, including RLHF methodology and instruction tuning concepts. Practiced evaluation and quality assessment approaches for LLM outputs to support AI response rating and red-teaming style tasks. Built familiarity with labeling-related guidance such as taxonomy design and inter-annotator agreement principles. • Studied LLM architectures, tokenization, instruction tuning, and RLHF via Google Cloud and IBM coursework. • Trained in identifying hallucinations, output quality issues, and bias in model responses. • Practiced writing unambiguous prompts and evaluating AI outputs for helpfulness, accuracy, and safety. • Reviewed responsible AI principles and labeling workflow concepts including annotation guidelines and agreement.