Recent research highlights the striking development of Large Language Models (LLMs) like ChatGPT, Flan-T5-Large, Vicuna, Llama 2, BLOOM, and GPT-4, which exhibit a remarkable ability to understand emotional cues. The study, led by Li et al., assessed LLMs’ emotional intelligence through various tasks and found that incorporating emotional prompts enhances their effectiveness.
Follow us on:
Read More News #latestmalaysia
AI Models and Emotions
This approach, named ‘EmotionPrompt,’ led to substantial performance improvements, with an 8% relative boost in instruction induction and a remarkable 115% increase in BIG-Bench tests. Additionally, a human study demonstrated a 10.9% average enhancement in generative task performance when using EmotionPrompt. These findings indicate the evolving capabilities of AI models, empowering users to optimize responses in novel ways.
Browse


