AI Annotator (DataAnnotation)
Text
RLHFServed as an AI Annotator using RLHF-related workflows to improve model response accuracy and learning efficiency. Designed and executed prompt-engineering strategies to test and refine LLM behaviors across multiple parameters. Performed safety- and quality-focused evaluation by assessing instruction-following, logical reasoning, and linguistic nuance. • Ranked and provided qualitative feedback on competing model outputs to refine model ranking algorithms. • Conducted copyediting and fact-checking to reduce hallucinations and maintain data integrity. • Ensured responses adhered to legal and ethical obligations, emphasizing harmlessness, truthfulness, and safety protocols. • Produced detailed assessments of model performance across multiple evaluation dimensions.