Senior AI Trainer (AI output evaluation and RLHF ranking)
Evaluated AI-generated outputs for factual accuracy, logical reasoning, safety, and instruction adherence to ensure high-quality deliverables. Performed RLHF preference and ranking assessments using pairwise and rubric-based methods to support model improvement. • Factual accuracy and reasoning checks • Safety and policy adherence evaluation • Instruction-following quality assessment • Output ranking via pairwise/rubric RLHF