Outlier
Annotate conversations between users and an AI assistant, judging whether the assistant's final response correctly matches the language, script, and dialect expected by the user — all crucial for training a granular Language Identification (LID) model.