Content Safety Evaluation Project
I performed labeling of harmful and sensitive content in text data for content safety evaluation purposes. My role included the identification and categorization of hate, misinformation, and unsafe advice within model responses or user-submitted texts. I meticulously documented the logic behind difficult or nuanced edge case decisions per project guidelines. • Labeled data for inappropriate, harmful, or unsafe content categories. • Handled complex edge cases requiring judgment and documentation. • Worked with content safety guidelines and quality assurance processes. • Collaborated with a team to maintain high labeling accuracy and reliability.