Measuring Common Sense
Yejin explains the challenges of measuring common sense reasoning, highlighting the limitations of multiple choice questions and the need for more generative evaluation methods. She also discusses the importance of human evaluation and crowdsourcing for gaining insights into what works in quantifying common sense.In this clip
From this podcast
The Gradient
Yejin Choi: Teaching Machines Common Sense and Morality
Related Questions
What challenges are faced in evaluating generative AI systems as highlighted within the discussions about model outputs and metrics in the episode Holistic Evaluation of Generative AI Systems // Jineet Doshi // #280 and the clip Evaluating Generative AI?
What challenges are faced in evaluating generative AI systems as highlighted within the discussions about model outputs and metrics in the episode Holistic Evaluation of Generative AI Systems // Jineet Doshi // #280 and the clip Evaluating Generative AI?