Look Within, Why LLMs Hallucinate: A Causal Perspective

• URL: http://arxiv.org/abs/2407.10153v1
• Date: Sun, 14 Jul 2024 10:47:44 GMT
• Title: Look Within, Why LLMs Hallucinate: A Causal Perspective
• Authors: He Li, Haoang Chi, Mingyu Liu, Wenjing Yang,
• Abstract summary: Large language models (LLMs) are a milestone in generative artificial intelligence, achieving significant success in text comprehension and generation tasks. LLMs suffer from severe hallucination problems, posing significant challenges to the practical applications of LLMs. We propose a method to intervene in LLMs' self-attention layers and maintain their structures and sizes intact.
• Score: 16.874588396996764
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: The emergence of large language models (LLMs) is a milestone in generative artificial intelligence, achieving significant success in text comprehension and generation tasks. Despite the tremendous success of LLMs in many downstream tasks, they suffer from severe hallucination problems, posing significant challenges to the practical applications of LLMs. Most of the works about LLMs' hallucinations focus on data quality. Self-attention is a core module in transformer-based LLMs, while its potential relationship with LLMs' hallucination has been hardly investigated. To fill this gap, we study this problem from a causal perspective. We propose a method to intervene in LLMs' self-attention layers and maintain their structures and sizes intact. Specifically, we disable different self-attention layers in several popular open-source LLMs and then compare their degrees of hallucination with the original ones. We evaluate the intervened LLMs on hallucination assessment benchmarks and conclude that disabling some specific self-attention layers in the front or tail of the LLMs can alleviate hallucination issues. The study paves a new way for understanding and mitigating LLMs' hallucinations.

Related papers

• MedHalu: Hallucinations in Responses to Healthcare Queries by Large Language Models [26.464489158584463]
  We conduct a pioneering study of hallucinations in LLM-generated responses to real-world healthcare queries from patients. We propose MedHalu, a carefully crafted first-of-its-kind medical hallucination dataset with a diverse range of health-related topics. We also introduce MedHaluDetect framework to evaluate capabilities of various LLMs in detecting hallucinations.
  arXiv  Detail & Related papers  (2024-09-29T00:09:01Z)
• LLM Internal States Reveal Hallucination Risk Faced With a Query [62.29558761326031]
  Humans have a self-awareness process that allows us to recognize what we don't know when faced with queries. This paper investigates whether Large Language Models can estimate their own hallucination risk before response generation. By a probing estimator, we leverage LLM self-assessment, achieving an average hallucination estimation accuracy of 84.32% at run time.
  arXiv  Detail & Related papers  (2024-07-03T17:08:52Z)
• Does Object Grounding Really Reduce Hallucination of Large Vision-Language Models? [53.89380284760555]
  Large vision-language models (LVLMs) produce captions that mention concepts that cannot be found in the image. These hallucinations erode the trustworthiness of LVLMs and are arguably among the main obstacles to their ubiquitous adoption. Recent work suggests that addition of grounding objectives -- those that explicitly align image regions or objects to text spans -- reduces the amount of LVLM hallucination.
  arXiv  Detail & Related papers  (2024-06-20T16:56:11Z)
• Do LLMs Know about Hallucination? An Empirical Investigation of LLM's Hidden States [19.343629282494774]
  Large Language Models (LLMs) can make up answers that are not real, and this is known as hallucination. This research aims to see if, how, and to what extent LLMs are aware of hallucination.
  arXiv  Detail & Related papers  (2024-02-15T06:14:55Z)
• The Dawn After the Dark: An Empirical Study on Factuality Hallucination in Large Language Models [134.6697160940223]
  hallucination poses great challenge to trustworthy and reliable deployment of large language models. Three key questions should be well studied: how to detect hallucinations (detection), why do LLMs hallucinate (source), and what can be done to mitigate them. This work presents a systematic empirical study on LLM hallucination, focused on the the three aspects of hallucination detection, source and mitigation.
  arXiv  Detail & Related papers  (2024-01-06T12:40:45Z)
• Siren's Song in the AI Ocean: A Survey on Hallucination in Large Language Models [116.01843550398183]
  Large language models (LLMs) have demonstrated remarkable capabilities across a range of downstream tasks. LLMs occasionally generate content that diverges from the user input, contradicts previously generated context, or misaligns with established world knowledge.
  arXiv  Detail & Related papers  (2023-09-03T16:56:48Z)
• Evaluation and Analysis of Hallucination in Large Vision-Language Models [49.19829480199372]
  Large Vision-Language Models (LVLMs) have recently achieved remarkable success. LVLMs are still plagued by the hallucination problem. Hallucination refers to the information of LVLMs' responses that does not exist in the visual input.
  arXiv  Detail & Related papers  (2023-08-29T08:51:24Z)
• Halo: Estimation and Reduction of Hallucinations in Open-Source Weak Large Language Models [11.497989461290793]
  Large Language Models (LLMs) have revolutionized Natural Language Processing (NLP) Open-source LLMs with fewer parameters often suffer from severe hallucinations compared to their larger counterparts. This paper focuses on measuring and reducing hallucinations in BLOOM 7B, a representative of such weaker open-source LLMs.
  arXiv  Detail & Related papers  (2023-08-22T20:12:49Z)
• Evaluating Object Hallucination in Large Vision-Language Models [122.40337582958453]
  This work presents the first systematic study on object hallucination of large vision-language models (LVLMs) We find that LVLMs tend to generate objects that are inconsistent with the target images in the descriptions. We propose a polling-based query method called POPE to evaluate the object hallucination.
  arXiv  Detail & Related papers  (2023-05-17T16:34:01Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.