Freelance AI Data Annotator
I evaluated and fact-checked AI-generated content against rigorous guidelines, ensuring output accuracy, safety, and logical consistency. My work contributed directly to the improvement of large language model (LLM) performance by providing high-quality training data. I followed detailed protocols to assess model outputs in English and Japanese. • Applied established evaluation rubrics to LLM-generated responses • Used bilingual skills to review and annotate both English and Japanese text • Ensured compliance with safety, truthfulness, and logical accuracy standards • Provided structured feedback for continuous model refinement