Coherence and Fluency Quiz

What you'll be asked

1. Which of the following best describes 'coherence' in the context of a GenAI-generated response?
2. You notice your model is producing grammatically perfect sentences, but the overall argument contradicts itself halfway through. Which aspect of quality optimization should you address?
3. When fine-tuning a model to improve fluency, which of the following techniques is most effective?
4. A model generates text that is technically correct but feels robotic and repetitive. Which parameter adjustment is most likely to improve the perceived fluency?
5. In a long-form generation task, the model begins to 'hallucinate' or lose the thread of the original prompt, negatively impacting coherence. What is the most robust architectural approach to mitigate this during inference?

Answers and explanations come after you submit.