mlsanityData to 14 Oct 2026

AggTruth: Contextual Hallucination Detection using Aggregated Attention Scores in LLMs

arXiv 2506.18628 ↗First posted 23 Jun 2025cs.AI cs.CLcited by 2

On the map: Retrieval-augmented generation → Hallucination detection in RAG

Abstract (arXiv)

In real-world applications, Large Language Models (LLMs) often hallucinate, even in Retrieval-Augmented Generation (RAG) settings, which poses a significant challenge to their deployment. In this paper, we introduce AggTruth, a method for online detection of contextual hallucinations by analyzing the distribution of internal attention scores in the provided context (passage). Specifically, we propose four different variants of the method, each varying in the aggregation technique used to calculate attention scores. Across all LLMs examined, AggTruth demonstrated stable performance in both same-task and cross-task setups, outperforming the current SOTA in multiple scenarios. Furthermore, we conducted an in-depth analysis of feature selection techniques and examined how the number of selected attention heads impacts detection performance, demonstrating that careful selection of heads is essential to achieve optimal results.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Placed in Hallucination detection in RAG, one of 15 directions in Retrieval-augmented generation. Builds on 15 of its 18 references that are in the map. First posted on arXiv 23 Jun 2025.

Builds on

15 of 18 references.

References · 15

  1. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,297
  2. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,754
  3. 2022Survey of Hallucination in Natural Language Generationcited by 2,075
  4. 2024Gemma 2: Improving Open Language Models at a Practical Sizecited by 1,998
  5. 2023A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questionscited by 1,601
  6. 2018Don't Give Me the Details, Just the Summary! Topic-Aware Convolutional Neural Networks for Extreme Summarizationcited by 1,317
  7. 2017Get To The Point: Summarization with Pointer-Generator Networkscited by 1,292
  8. 2023RWKV: Reinventing RNNs for the Transformer Eracited by 827
  9. 2023SelfCheckGPT: Zero-Resource Black-Box Hallucination Detection for Generative Large Language Modelscited by 810
  10. 2023The Internal State of an LLM Knows When It's Lyingcited by 659
  11. 2024Does Fine-Tuning LLMs on New Knowledge Encourage Hallucinations?cited by 232
  12. 2023ChatGPT: Jack of all trades, master of nonecited by 204
3 more
  1. 2024Lookback Lens: Detecting and Mitigating Contextual Hallucinations in Large Language Models Using Only Attention Mapscited by 142
  2. 2022Continual Training of Language Models for Few-Shot Learningcited by 33
  3. 2024Into the Unknown: Self-Learning Large Language Modelscited by 5

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026Gradient-Stable Attention Heads Signal LLM Correctnesscited by 5
  2. 2026What properties of reasoning supervision are associated with improved downstream model quality?cited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Attention-guided Self-reflection for Zero-shot Hallucination Detection in Large Language Models
  2. 2026RAGognizer: Hallucination-Aware Fine-Tuning via Detection Head Integration
  3. 2023RAGTruth: A Hallucination Corpus for Developing Trustworthy Retrieval-Augmented Language Models
  4. 2025LUMINA: Detecting Hallucinations in RAG System with Context-Knowledge Signals
  5. 2025InterpDetect: Interpretable Signals for Detecting Hallucinations in Retrieval-Augmented Generation
  6. 2025Bi'an: A Bilingual Benchmark and Model for Hallucination Detection in Retrieval-Augmented Generation
  7. 2025Osiris: A Lightweight Open-Source Hallucination Detection System
  8. 2025ORION Grounded in Context: Retrieval-Based Method for Hallucination Detection