Automatic Generation of Titles for Research Papers Using Language Models
Researchers have developed a method for automatically generating research paper titles from abstracts using large language models. They curated a new dataset, SpringerSSAT, and evaluated several models including PEGASUS-large, LLaMA-3-8B, and GPT-3.5-turbo. Experiments showed that a fine-tuned PEGASUS-large model performed best across multiple metrics, outperforming even GPT-3.5-turbo in a zero-shot setting. AI
IMPACT Automated title generation could streamline the academic publishing process and improve the discoverability of research.