AI Training Data Contributor — Outlier AI (Remote, Contract)
AI training data contribution supporting large language model improvement through written evaluation and response ranking. • Authored prompt-response pairs across creative writing, factual Q&A, and reasoning tasks. • Ranked model responses using rubrics for accuracy, relevance, and helpfulness. • Identified and reported hallucinations, factual errors, and formatting issues. • Provided feedback to support iterative training rounds and model refinement.