Sematic Search Augmented Conversation for Enhanced Dialogue Generation
Abstract
Although advanced conversational models like ChatGPT are capable of generating rich and coherent content, the generated responses often contain fictional facts and knowledge hallucinations.A mainstream approach to addressing this problem in the past has been fine-tuning or retraining models by injecting external knowledge into pre-trained language models.However, given the enormous scale of current state-of-the-art language models, these methods require continuous retraining to update the knowledge embedded in the model parameters, which is undeniably challenging, slow, expensive, and the updated models lack scalability.In this work, we explore the use of semantic search based on user input and local knowledge to prompt language models for enhanced dialogue generation.We experiment with different domains of dialogue on four popular large language models (LLMs), and the results show that our approach, compared to the method of injecting knowledge into LLMs, can effectively improve the utilization efficiency of knowledge, significantly reduce knowledge hallucination problems, and has almost unlimited scalability.
Read the paper
Related studies
- Knowledge Management Perspective of Generative Artificial Intelligence
- Navigating Generative Artificial Intelligence Promises and Perils for Knowledge and Creative Work
- Characterizing the evolving research on enterprise content management
- Large language models and multimodal foundation models for precision oncology
- From Generative AI to Extended Reality: Multidisciplinary Perspectives on the Challenges, Opportunities, and Future of Educational Computing
- Generative Artificial Intelligence: A New Engine for Advancing Environmental Science and Engineering
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.