Lookback Lens: Detecting and Mitigating Contextual Hallucinations in Large Language Models Using Only Attention Maps

• URL: http://arxiv.org/abs/2407.07071v2
• Date: Thu, 3 Oct 2024 17:26:48 GMT
• Title: Lookback Lens: Detecting and Mitigating Contextual Hallucinations in Large Language Models Using Only Attention Maps
• Authors: Yung-Sung Chuang, Linlu Qiu, Cheng-Yu Hsieh, Ranjay Krishna, Yoon Kim, James Glass,
• Abstract summary: Large language models (LLMs) can hallucinate details and respond with unsubstantiated answers. This paper describes a simple approach for detecting such contextual hallucinations.
• Score: 48.58310785625051
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: When asked to summarize articles or answer questions given a passage, large language models (LLMs) can hallucinate details and respond with unsubstantiated answers that are inaccurate with respect to the input context. This paper describes a simple approach for detecting such contextual hallucinations. We hypothesize that contextual hallucinations are related to the extent to which an LLM attends to information in the provided context versus its own generations. Based on this intuition, we propose a simple hallucination detection model whose input features are given by the ratio of attention weights on the context versus newly generated tokens (for each attention head). We find that a linear classifier based on these lookback ratio features is as effective as a richer detector that utilizes the entire hidden states of an LLM or a text-based entailment model. The lookback ratio-based detector -- Lookback Lens -- is found to transfer across tasks and even models, allowing a detector that is trained on a 7B model to be applied (without retraining) to a larger 13B model. We further apply this detector to mitigate contextual hallucinations, and find that a simple classifier-guided decoding approach is able to reduce the amount of hallucination, for example by 9.6% in the XSum summarization task.

Related papers

• MetaToken: Detecting Hallucination in Image Descriptions by Meta Classification [1.3654846342364308]
  We introduce MetaToken, a lightweight binary classifier to detect hallucinations on the token-level at negligible cost. Based on a statistical analysis, we reveal key factors of hallucinations in LVLMs which have been overseen in previous works. We evaluate our method on four state-of-the-art LVLMs demonstrating the effectiveness of our approach.
  arXiv  Detail & Related papers  (2024-05-29T15:28:42Z)
• Detecting and Mitigating Hallucination in Large Vision Language Models via Fine-Grained AI Feedback [48.065569871444275]
  We propose detecting and mitigating hallucinations in Large Vision Language Models (LVLMs) via fine-grained AI feedback. We generate a small-size hallucination annotation dataset by proprietary models. Then, we propose a detect-then-rewrite pipeline to automatically construct preference dataset for training hallucination mitigating model.
  arXiv  Detail & Related papers  (2024-04-22T14:46:10Z)
• Hallucination Augmented Contrastive Learning for Multimodal Large Language Model [53.65682783591723]
  Multi-modal large language models (MLLMs) have been shown to efficiently integrate natural language with visual information to handle multi-modal tasks. However, MLLMs still face a fundamental limitation of hallucinations, where they tend to generate erroneous or fabricated information. In this paper, we address hallucinations in MLLMs from a novel perspective of representation learning.
  arXiv  Detail & Related papers  (2023-12-12T04:05:15Z)
• DelucionQA: Detecting Hallucinations in Domain-specific Question Answering [22.23664008053246]
  Hallucination is a well-known phenomenon in text generated by large language models (LLMs) We introduce a dataset, DelucionQA, that captures hallucinations made by retrieval-augmented LLMs for a domain-specific QA task. We propose a set of hallucination detection methods to serve as baselines for future works from the research community.
  arXiv  Detail & Related papers  (2023-12-08T17:41:06Z)
• A New Benchmark and Reverse Validation Method for Passage-level Hallucination Detection [63.56136319976554]
  Large Language Models (LLMs) generate hallucinations, which can cause significant damage when deployed for mission-critical tasks. We propose a self-check approach based on reverse validation to detect factual errors automatically in a zero-resource fashion. We empirically evaluate our method and existing zero-resource detection methods on two datasets.
  arXiv  Detail & Related papers  (2023-10-10T10:14:59Z)
• AutoHall: Automated Hallucination Dataset Generation for Large Language Models [56.92068213969036]
  This paper introduces a method for automatically constructing model-specific hallucination datasets based on existing fact-checking datasets called AutoHall. We also propose a zero-resource and black-box hallucination detection method based on self-contradiction.
  arXiv  Detail & Related papers  (2023-09-30T05:20:02Z)
• Detecting and Preventing Hallucinations in Large Vision Language Models [4.7264116948935975]
  M-HalDetect is the first multi-modal hallucination detection dataset for detailed image descriptions. We train fine-grained multi-modal reward models from InstructBLIP and evaluate their effectiveness with best-of-n rejection sampling. We find that our reward model generalizes to other multi-modal models, reducing hallucinations in LLaVA and mPLUG-OWL by 15% and 57% respectively.
  arXiv  Detail & Related papers  (2023-08-11T21:35:20Z)
• Trapping LLM Hallucinations Using Tagged Context Prompts [11.655802601887197]
  We propose a novel method to recognize and flag instances when large language models perform outside their domain knowledge. We find that the use of context combined with embedded tags can successfully combat hallucinations within generative language models.
  arXiv  Detail & Related papers  (2023-06-09T17:48:54Z)
• SelfCheckGPT: Zero-Resource Black-Box Hallucination Detection for Generative Large Language Models [55.60306377044225]
  "SelfCheckGPT" is a simple sampling-based approach to fact-check the responses of black-box models. We investigate this approach by using GPT-3 to generate passages about individuals from the WikiBio dataset.
  arXiv  Detail & Related papers  (2023-03-15T19:31:21Z)

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.