Unsupervised Hallucination Detection by Inspecting Reasoning Processes
Unsupervised hallucination detection aims to identify hallucinated content generated by large language models (LLMs) without relying on labeled data. While unsupervised methods have gained popularity by eliminating labor-intensive human annotations, they frequently rely on proxy signals unrelated to factual correctness. This misalignment biases detection probes toward superficial or non-truth-related aspects, limiting generalizability across datasets and scenarios. To overcome these limitations, we propose IRIS, an unsupervised hallucination detection framework, leveraging internal representations intrinsic to factual correctness. IRIS prompts the LLM to carefully verify the truthfulness of a given statement, and obtain its contextualized embedding as informative features for training. Meanwhile, the uncertainty of each response is considered a soft pseudolabel for truthfulness. Experimental results demonstrate that IRIS consistently outperforms existing unsupervised methods. Our approach is fully unsupervised, computationally low cost, and works well even with few training data, making it suitable for real-time detection.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Leveraging Low-Level Symbolic Competences for Unsupervised Grounding in Hallucination Detection
Hallucination-where a language model generates outputs that are factually incorrect or unsupported by the source-is a major challenge for both prompted and fine-tuned language models. Detecting hallucinations is difficul…
HARP: Hallucination Detection via Reasoning Subspace Projection
Hallucinations in Large Language Models (LLMs) pose a major barrier to their reliable use in critical decision-making. Although existing hallucination detection methods have improved accuracy, they still struggle with di…
Detecting Hallucinations in Retrieval-Augmented Generation via Semantic-level Internal Reasoning Graph
The Retrieval-augmented generation (RAG) system based on Large language model (LLM) has made significant progress. It can effectively reduce factuality hallucinations, but faithfulness hallucinations still exist. Previou…
Hallucinated but Factual! Inspecting the Factuality of Hallucinations in Abstractive Summarization
State-of-the-art abstractive summarization systems often generate hallucinations; i.e., content that is not directly inferable from the source text. Despite being assumed to be incorrect, we find that much hallucinated c…
Abstractive Text SummarizationReinforcement Learning (RL)World KnowledgeHallucinated but Factual! Inspecting the Factuality of Hallucinations in Abstractive Summarization
State-of-the-art abstractive summarization systems often generate \emph{hallucinations}; i.e., content that is not directly inferable from the source text. Despite being assumed incorrect, we find that much hallucinated …
Abstractive Text SummarizationReinforcement Learning (RL)World Knowledge