HaloScope: Harnessing Unlabeled LLM Generations for Hallucination Detection
원문 초록
The surge in applications of large language models (LLMs) has prompted concerns about the generation of misleading or fabricated information, known as hallucinations. Therefore, detecting hallucinations has become critical to maintaining trust in LLM-generated content. A primary challenge in learning a truthfulness classifier is the lack of a large amount of labeled truthful and hallucinated data. To address the challenge, we introduce HaloScope, a novel learning framework that leverages the unlabeled LLM generations in the wild for hallucination detection. Such unlabeled data arises freely upon deploying LLMs in the open world, and consists of both truthful and hallucinated information. To harness the unlabeled data, we present an automated membership estimation score for distinguishing between truthful and untruthful generations within unlabeled mixture data, thereby enabling the training of a binary truthfulness classifier on top. Importantly, our framework does not require extra data collection and human annotations, offering strong flexibility and practicality for real-world applications. Extensive experiments show that HaloScope can achieve superior hallucination detection performance, outperforming the competitive rivals by a significant margin. Code is available at https://github.com/deeplearningwisc/haloscope.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
Mitigating Object Hallucinations via Sentence-Level Early Intervention
UQLM: A Python Package for Uncertainty Quantification in Large Language Models
Mitigating Hallucination of Large Vision-Language Models via Dynamic Logits Calibration
KnowRL: Exploring Knowledgeable Reinforcement Learning for Factuality
DiscoSG: Towards Discourse-Level Text Scene Graph Parsing through Iterative Graph Refinement
Abstract Meaning Representation for Hospital Discharge Summarization
토론
이 논문과 연결된 포럼 글이 아직 없습니다.