mlsanityData to 7 Oct 2026

Hallucination Detection in LLMs with Topological Divergence on Attention Graphs

arXiv 2504.10063 ↗First posted 14 Apr 2025cs.CL cs.AI math.ATcited by 18

Not on the map. Why?

Abstract (arXiv)

Hallucination, i.e., generating factually incorrect content, remains a critical challenge for large language models (LLMs). We introduce TOHA, a TOpology-based HAllucination detector in the RAG setting, which leverages a topological divergence metric to quantify the structural properties of graphs induced by attention matrices. Examining the topological divergence between prompt and response subgraphs reveals consistent patterns: higher divergence values in specific attention heads correlate with hallucinated outputs, independent of the dataset. Extensive experiments - including evaluation on question answering and summarization tasks - show that our approach achieves state-of-the-art or competitive results on several benchmarks while requiring minimal annotated data and computational resources. Our findings suggest that analyzing the topological structure of attention matrices can serve as an efficient and robust indicator of factual reliability in LLMs.

Key facts. Cited by 18 papers in the map (data through 7 Oct 2026). Builds on 40 of its 71 references that are in the map. First posted on arXiv 14 Apr 2025.

Builds on

40 of 71 references.

References · 40

  1. 2017Attention Is All You Needcited by 59,932
  2. 2020Retrieval-Augmented Generation for Knowledge-Intensive NLP Taskscited by 9,160
  3. 2016SQuAD: 100,000+ Questions for Machine Comprehension of Textcited by 4,115
  4. 2018HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answeringcited by 3,686
  5. 2016MS MARCO: A Human Generated MAchine Reading COmprehension Datasetcited by 1,591
  6. 2018Don't Give Me the Details, Just the Summary! Topic-Aware Convolutional Neural Networks for Extreme Summarizationcited by 1,308
  7. 2019Analyzing Multi-Head Self-Attention: Specialized Heads Do the Heavy Lifting, the Rest Can Be Prunedcited by 844
  8. 2016Abstractive Text Summarization Using Sequence-to-Sequence RNNs and Beyondcited by 809
  9. 2018CoQA: A Conversational Question Answering Challengecited by 674
  10. 2023The Internal State of an LLM Knows When It's Lyingcited by 644
  11. 2018Deep learning for time series classification: a reviewcited by 575
  12. 2023Trusting Your Evidence: Hallucinate Less with Context-aware Decodingcited by 319
28 more
  1. 2024INSIDE: LLMs' Internal States Retain the Power of Hallucination Detectioncited by 317
  2. 2023RAGTruth: A Hallucination Corpus for Developing Trustworthy Retrieval-Augmented Language Modelscited by 272
  3. 2019Analyzing the Structure of Attention in a Transformer Language Modelcited by 241
  4. 2024LLMs instead of Human Judges? A Large Scale Empirical Study across 20 NLP Evaluation Taskscited by 236
  5. 2024LLMs Know More Than They Show: On the Intrinsic Representation of LLM Hallucinationscited by 230
  6. 2022Out-of-Distribution Detection and Selective Generation for Conditional Language Modelscited by 193
  7. 2017An introduction to Topological Data Analysis: fundamental and practical aspects for data scientistscited by 171
  8. 2024Fact-Checking the Output of Large Language Models via Token-Level Uncertainty Quantificationcited by 159
  9. 2024Kernel Language Entropy: Fine-grained Uncertainty Quantification for LLMs from Semantic Similaritiescited by 156
  10. 2024Lookback Lens: Detecting and Mitigating Contextual Hallucinations in Large Language Models Using Only Attention Mapscited by 140
  11. 2021Hallucinated but Factual! Inspecting the Factuality of Hallucinations in Abstractive Summarizationcited by 128
  12. 2025Why do LLMs attend to the first token?cited by 122
  13. 2023Intrinsic Dimension Estimation for Robust Detection of AI-Generated Textscited by 111
  14. 2024ReDeEP: Detecting Hallucination in Retrieval-Augmented Generation via Mechanistic Interpretabilitycited by 100
  15. 2024A Comprehensive Survey of Hallucination in Large Language, Image, Video and Audio Foundation Modelscited by 98
  16. 2023Successor Heads: Recurring, Interpretable Attention Heads In The Wildcited by 77
  17. 2023Do Androids Know They're Only Dreaming of Electric Sheep?cited by 67
  18. 2021Artificial Text Detection via Examining the Topology of Attention Mapscited by 49
  19. 2025Hallucination Detection in LLMs Using Spectral Features of Attention Mapscited by 42
  20. 2021Manifold Topology Divergence: a Framework for Comparing Data Manifoldscited by 31
  21. 2020GPU-Accelerated Computation of Vietoris-Rips Persistence Barcodescited by 26
  22. 2025The Dual-Route Model of Inductioncited by 22
  23. 2024A Probabilistic Framework for LLM Hallucination Detection via Belief Tree Propagationcited by 21
  24. 2022Acceptability Judgements via Examining the Topology of Attention Mapscited by 19
  25. 2021On the Distribution, Sparsity, and Inference-time Quantization of Attention Values in Transformerscited by 14
  26. 2022Topological Data Analysis for Speech Processingcited by 9
  27. 2023Uncertainty Estimation of Transformers' Predictions via Topological Analysis of the Attention Matricescited by 8
  28. 2023Can BERT eat RuCoLA? Topological Data Analysis to Explaincited by 5

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 4; 2026: 14.

18 papers.

Sort:

By area:

  1. 2025Large Language Models Hallucination: A Comprehensive Surveycited by 45
  2. 2025Shadows in the Attention: Contextual Perturbation and Representation Drift in the Dynamics of Hallucination in LLMscited by 9
  3. 2026HALT: Hallucination Assessment via Log-probs as Time seriescited by 8
  4. 2025Neural Message-Passing on Attention Graphs for Hallucination Detectioncited by 5
  5. 2026Student Guides Teacher: Weak-to-Strong Inference via Spectral Orthogonal Explorationcited by 3
  6. 2026HalluZig: Hallucination Detection using Zigzag Persistencecited by 2
  7. 2025Probabilistic distances-based hallucination detection in LLMs with RAGcited by 1
  8. 2026Pre-Generation Hallucination Detection in Large Language Models via Soft-Target Attention Probingcited by 1
  9. 2026Enhancing the Reliability of Medical AI through Expert-guided Uncertainty Modelingcited by 0
  10. 2026Uncertainty Estimation for the Open-Set Text Classification systemscited by 0
  11. 2026Attention Sinks as Internal Signals for Hallucination Detection in Large Language Modelscited by 0
  12. 2026Topology-Aware Layer Pruning for Large Vision-Language Modelscited by 0
6 more
  1. 2026Self-Attention as Transport: Limits of Symmetric Spectral Diagnosticscited by 0
  2. 2026ASK-NN: An Asymmetric Nearest-Neighbor Test that detects Distribution Drifts in Natural Languagecited by 0
  3. 2026SIRIN: A Unified Toolkit for Detecting Contextual Hallucinations in Retrieval-Augmented and Memory-Grounded LLM Systemscited by 0
  4. 2026CodeTD: Topology of Attention Detects Hallucinations in Code LLMscited by 0
  5. 2026Detecting Hallucination in LLMs: Tracing the Topological Signatures of Impaired Context Sharingcited by 0
  6. 2026Mixed-Prior Decision Risk for Open-Set Recognitioncited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025HalluField: Detecting LLM Hallucinations via Field-Theoretic Modeling
  2. 2026Evidence Graph Consistency in Retrieval-Augmented Generation: A Model-Dependent Analysis of Hallucination Detection
  3. 2026TDGNet: Hallucination Detection in Diffusion Language Models via Temporal Dynamic Graphs
  4. 2023A Survey on Hallucination in Large Language Models: Principles, Taxonomy, Challenges, and Open Questions
  5. 2026Detecting Hallucinations for Large Language Model-based Knowledge Graph Reasoning
  6. 2026Unmasking Hallucinations: A Causal Graph-Attention Perspective on Factual Reliability in Large Language Models
  7. 2025D2HScore: Reasoning-Aware Hallucination Detection via Semantic Breadth and Depth Analysis in LLMs
  8. 2025SelfCheck-Eval: A Multi-Module Framework for Zero-Resource Hallucination Detection in Large Language Models