Poetry Generation

Automatic poetry generation aims to create poems using computational methods, primarily focusing on achieving both high-quality output and stylistic diversity. Current research heavily utilizes large language models (LLMs), often fine-tuned on large poetry corpora or employing techniques like multi-agent systems and diffusion models to enhance control over semantic and metrical aspects. This field contributes to natural language generation research by providing a challenging testbed for evaluating model capabilities in creative tasks, while also offering potential applications in literary studies and creative writing tools.