Synthetic Dialogue Generation with Role-Playing LLMs II - Day 9 Afternoon

Explore advanced techniques for generating synthetic dialogues using role-playing large language models in this comprehensive tutorial continuation from JSALT 2025. Delve into sophisticated methodologies presented by leading researchers from University of Toulon, Idiap Research Institute, Johns Hopkins University, University of Zaragoza, and Harvard University. Master the intricacies of leveraging LLMs to create realistic conversational scenarios through role-playing mechanisms, building upon foundational concepts to tackle more complex dialogue generation challenges. Examine cutting-edge approaches to synthetic data creation, role assignment strategies, and dialogue quality assessment techniques. Learn how to implement and optimize role-playing frameworks for various dialogue applications, from conversational AI systems to training data augmentation. Gain insights into the latest research developments in computational linguistics and natural language processing as they apply to dialogue systems and synthetic conversation generation.