Emotionally Numb or Empathetic? Evaluating How LLMs Feel Using EmotionBench
This work addresses the need for better evaluation of anthropomorphic capabilities in LLMs, particularly for applications in human-AI interaction, but it is incremental as it builds on existing emotion appraisal theory and benchmarking approaches.
The authors tackled the problem of evaluating empathy in Large Language Models (LLMs) by creating EmotionBench, a dataset of over 400 situations to test emotional responses, and found that while LLMs can respond appropriately in some cases, they generally misalign with human emotional behaviors and fail to connect similar situations.
Evaluating Large Language Models' (LLMs) anthropomorphic capabilities has become increasingly important in contemporary discourse. Utilizing the emotion appraisal theory from psychology, we propose to evaluate the empathy ability of LLMs, i.e., how their feelings change when presented with specific situations. After a careful and comprehensive survey, we collect a dataset containing over 400 situations that have proven effective in eliciting the eight emotions central to our study. Categorizing the situations into 36 factors, we conduct a human evaluation involving more than 1,200 subjects worldwide. With the human evaluation results as references, our evaluation includes seven LLMs, covering both commercial and open-source models, including variations in model sizes, featuring the latest iterations, such as GPT-4, Mixtral-8x22B, and LLaMA-3.1. We find that, despite several misalignments, LLMs can generally respond appropriately to certain situations. Nevertheless, they fall short in alignment with the emotional behaviors of human beings and cannot establish connections between similar situations. Our collected dataset of situations, the human evaluation results, and the code of our testing framework, i.e., EmotionBench, are publicly available at https://github.com/CUHK-ARISE/EmotionBench.