WikiChat: Stopping the Hallucination of Large Language Model Chatbots by Few-Shot Grounding on Wikipedia

Abstract

This paper presents the first few-shot LLM-based chatbot that almost never hallucinates and has high conversationality and low latency. WikiChat is grounded on the English Wikipedia, the largest curated free-text corpus. WikiChat generates a response from an LLM, retains only the grounded facts, and combines them with additional information it retrieves from the corpus to form factual and engaging responses. We distill WikiChat based on GPT-4 into a 7B-parameter LLaMA model with minimal loss of quality, to significantly improve its latency, cost and privacy, and facilitate research and deployment. Using a novel hybrid human-and-LLM evaluation methodology, we show that our best system achieves 97.3% factual accuracy in simulated conversations. It significantly outperforms all retrieval-based and LLM-based baselines, and by 3.9%, 38.6% and 51.0% on head, tail and recent knowledge compared to GPT-4. Compared to previous state-of-the-art retrieval-based chatbots, WikiChat is also significantly more informative and engaging, just like an LLM. WikiChat achieves 97.9% factual accuracy in conversations with human users about recent topics, 55.0% better than GPT-4, while receiving significantly higher user ratings and more favorable comments.

Findings of EMNLP 2023

References in corpus (10)

• LLaMA: Open and Efficient Foundation Language Models
• REALM: Retrieval-Augmented Language Model Pre-Training
• FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness
• A Multitask, Multilingual, Multimodal Evaluation of ChatGPT on Reasoning, Hallucination, and Interactivity
• Self-Refine: Iterative Refinement with Self-Feedback
• Atlas: Few-shot Learning with Retrieval Augmented Language Models
• Large Language Models Are State-of-the-Art Evaluators of Translation Quality
• A Survey on Retrieval-Augmented Text Generation
• AnnoLLM: Making Large Language Models to Be Better Crowdsourced Annotators
• Memory-Based Model Editing at Scale

Cited by in corpus (4)

• "It's like a rubber duck that talks back": Understanding Generative AI-Assisted Data Analysis Workflows through a Participatory Prompting Study
• WebANNS: Fast and Efficient Approximate Nearest Neighbor Search in Web Browsers
• GRILLBot In Practice: Lessons and Tradeoffs Deploying Large Language Models for Adaptable Conversational Task Assistants
• Data Work in Memory Institutions: Why and How Information Professionals Use Wikidata