COCOA and EVAL Project (french)
The Cocoa Project on Remotasks mainly involves text-based labeling related to LLM training, and the main labeling types include: 🔹 Ranking / Comparison (Pairwise or Multi-choice) You are shown two or more AI-generated responses to the same prompt. Your task is to rank them from best to worst based on helpfulness, clarity, relevance, and safety. 🔹 Scoring / Rating You give scores (usually from 1 to 5) on specific criteria like: Helpfulness Accuracy Toxicity / Bias Clarity 🔹 Rewrite / Generation You are asked to edit or rewrite an AI response to make it better. Sometimes, you write a new response from scratch based on a user prompt.