Mitigating Entity-Level Hallucination in Large Language Models

The emergence of Large Language Models (LLMs) has revolutionized how users access information, shifting from traditional search engines to direct question-and-answer interactions with LLMs. However, the widespread adoption of LLMs has revealed a significant challenge known as hallucination, wherein LLMs generate coherent yet factually inaccurate responses. This hallucination phenomenon has led to users' distrust in information retrieval systems based on LLMs. To tackle this challenge, this paper proposes Dynamic Retrieval Augmentation based on hallucination Detection (DRAD) as a novel method to detect and mitigate hallucinations in LLMs. DRAD improves upon traditional retrieval augmentation by dynamically adapting the retrieval process based on real-time hallucination detection. It features two main components: Real-time Hallucination Detection (RHD) for identifying potential hallucinations without external models, and Self-correction based on External Knowledge (SEK) for correcting these errors using external knowledge. Experiment results show that DRAD demonstrates superior performance in both detecting and mitigating hallucinations in LLMs. All of our code and data are open-sourced at https://github.com/oneal2000/EntityHallucination.

DRAGIN: Dynamic Retrieval Augmented Generation based on the Real-time Information Needs of Large Language Models

Dynamic retrieval augmented generation (RAG) paradigm actively decides when and what to retrieve during the text generation process of Large Language Models (LLMs). There are two key elements of this paradigm: identifying the optimal moment to activate the retrieval module (deciding when to retrieve) and crafting the appropriate query once retrieval is triggered (determining what to retrieve). However, current dynamic RAG methods fall short in both aspects. Firstly, the strategies for deciding when to retrieve often rely on static rules. Moreover, the strategies for deciding what to retrieve typically limit themselves to the LLM's most recent sentence or the last few tokens, while the LLM's real-time information needs may span across the entire context. To overcome these limitations, we introduce a new framework, DRAGIN, i.e., Dynamic Retrieval Augmented Generation based on the real-time Information Needs of LLMs. Our framework is specifically designed to make decisions on when and what to retrieve based on the LLM's real-time information needs during the text generation process. We evaluate DRAGIN along with existing methods comprehensively over 4 knowledge-intensive generation datasets. Experimental results show that DRAGIN achieves superior performance on all tasks, demonstrating the effectiveness of our method. We have open-sourced all the code, data, and models in GitHub: https://github.com/oneal2000/DRAGIN/tree/main

Towards the Classification of Error-Related Potentials using Riemannian Geometry

The error-related potential (ErrP) is an event-related potential (ERP) evoked by an experimental participant's recognition of an error during task performance. ErrPs, originally described by cognitive psychologists, have been adopted for use in brain-computer interfaces (BCIs) for the detection and correction of errors, and the online refinement of decoding algorithms. Riemannian geometry-based feature extraction and classification is a new approach to BCI which shows good performance in a range of experimental paradigms, but has yet to be applied to the classification of ErrPs. Here, we describe an experiment that elicited ErrPs in seven normal participants performing a visual discrimination task. Audio feedback was provided on each trial. We used multi-channel electroencephalogram (EEG) recordings to classify ErrPs (success/failure), comparing a Riemannian geometry-based method to a traditional approach that computes time-point features. Overall, the Riemannian approach outperformed the traditional approach (78.2% versus 75.9% accuracy, p < 0.05); this difference was statistically significant (p < 0.05) in three of seven participants. These results indicate that the Riemannian approach better captured the features from feedback-elicited ErrPs, and may have application in BCI for error detection and correction.