AI Response Rater & Reviewer
As an AI Response Rater & Reviewer for Outlier AI's Ather Project on the MultiMango Platform, I assessed AI-generated responses for accuracy, clarity, and relevance. My feedback directly helped refine and improve large language model behaviors. I also identified hallucinations and bias, ensuring strict compliance with project requirements. • Compared AI-generated text outputs for alignment with prompts. • Provided structured, constructive feedback to developers. • Detected and documented logical errors and model hallucinations. • Maintained a high standard of evaluation accuracy throughout.