SINdex: Semantic INconsistency Index for Hallucination Detection in LLMs
This work addresses the issue of factually incorrect outputs in LLMs, which is critical for reliable deployment across domains, though it is incremental as it builds on existing uncertainty-based methods.
The paper tackles the problem of hallucination detection in large language models by introducing SINdex, a semantic inconsistency index, which achieves up to 9.3% AUROC improvement over state-of-the-art methods on QA datasets.
Large language models (LLMs) are increasingly deployed across diverse domains, yet they are prone to generating factually incorrect outputs - commonly known as "hallucinations." Among existing mitigation strategies, uncertainty-based methods are particularly attractive due to their ease of implementation, independence from external data, and compatibility with standard LLMs. In this work, we introduce a novel and scalable uncertainty-based semantic clustering framework for automated hallucination detection. Our approach leverages sentence embeddings and hierarchical clustering alongside a newly proposed inconsistency measure, SINdex, to yield more homogeneous clusters and more accurate detection of hallucination phenomena across various LLMs. Evaluations on prominent open- and closed-book QA datasets demonstrate that our method achieves AUROC improvements of up to 9.3% over state-of-the-art techniques. Extensive ablation studies further validate the effectiveness of each component in our framework.