Pun Generation

Pun generation research focuses on understanding and replicating the human ability to create humorous wordplay, primarily using large language models (LLMs) and multimodal approaches incorporating visual cues. Current research explores various aspects, including pun recognition, generation, and explanation, often employing techniques like prompt engineering and pre-training on user behavior data to improve model performance. This work contributes to a broader understanding of humor, language processing, and multimodal reasoning, with potential applications in creative writing, chatbot development, and even social media bot detection.