mlsanityData to 7 Oct 2026

Artificial Text Detection via Examining the Topology of Attention Maps

arXiv 2109.04825 ↗First posted 10 Sep 2021cs.CL cs.LG math.ATcited by 49

Not on the map. Why?

Abstract (arXiv)

The impressive capabilities of recent generative models to create texts that are challenging to distinguish from the human-written ones can be misused for generating fake news, product reviews, and even abusive content. Despite the prominent performance of existing methods for artificial text detection, they still lack interpretability and robustness towards unseen models. To this end, we propose three novel types of interpretable topological features for this task based on Topological Data Analysis (TDA) which is currently understudied in the field of NLP. We empirically show that the features derived from the BERT model outperform count- and neural-based baselines up to 10% on three common datasets, and tend to be the most robust towards unseen GPT-style generation models as opposed to existing methods. The probing analysis of the features reveals their sensitivity to the surface and syntactic properties. The results demonstrate that TDA is a promising line with respect to NLP tasks, specifically the ones that incorporate surface and structural information.

Key facts. Cited by 49 papers in the map (data through 7 Oct 2026). Builds on 34 of its 53 references that are in the map. First posted on arXiv 10 Sep 2021.

Builds on

34 of 53 references.

References · 34

  1. 2017Attention Is All You Needcited by 59,932
  2. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  3. 2020Language Models are Few-Shot Learnerscited by 32,199
  4. 2019RoBERTa: A Robustly Optimized BERT Pretraining Approachcited by 13,266
  5. 2019XLNet: Generalized Autoregressive Pretraining for Language Understandingcited by 2,801
  6. 2019What Does BERT Look At? An Analysis of BERT's Attentioncited by 994
  7. 2019Are Sixteen Heads Really Better than One?cited by 803
  8. 2019CTRL: A Conditional Transformer Language Model for Controllable Generationcited by 754
  9. 2019Defending Against Neural Fake Newscited by 553
  10. 2019Release Strategies and the Social Impacts of Language Modelscited by 505
  11. 2019GLTR: Statistical Detection and Visualization of Generated Textcited by 490
  12. 2018What you can cram into a single vector: Probing sentence embeddings for linguistic propertiescited by 452
22 more
  1. 2019Automatic Detection of Generated Text is Easiest when Humans are Fooledcited by 260
  2. 2016Diachronic Word Embeddings Reveal Statistical Laws of Semantic Changecited by 219
  3. 2017An introduction to Topological Data Analysis: fundamental and practical aspects for data scientistscited by 171
  4. 2020Automatic Detection of Machine Generated Text: A Critical Surveycited by 160
  5. 2019A Topological Loss Function for Deep-Learning based Image Segmentation using Persistent Homologycited by 134
  6. 2020TweepFake: about Detecting Deepfake Tweetscited by 117
  7. 2019Ripser: efficient computation of Vietoris-Rips persistence barcodescited by 115
  8. 2020Topology of deep neural networkscited by 111
  9. 2019Real or Fake? Learning to Discriminate Machine from Human Generated Textcited by 98
  10. 2019Do Attention Heads in BERT Track Syntactic Dependencies?cited by 87
  11. 2019A Topology Layer for Machine Learningcited by 79
  12. 2020The Radicalization Risks of GPT-3 and Advanced Neural Language Modelscited by 76
  13. 2018Neural Persistence: A Complexity Measure for Deep Neural Networks Using Algebraic Topologycited by 73
  14. 2020Computing the Testing Error without a Testing Setcited by 54
  15. 2019Generating Sentiment-Preserving Fake Online Reviews Using Neural Language Models and Their Human- and Machine-based Detectioncited by 50
  16. 2020GPU-Accelerated Computation of Vietoris-Rips Persistence Barcodescited by 26
  17. 2020How Furiously Can Colourless Green Ideas Sleep? Sentence Acceptability in Contextcited by 26
  18. 2020A Fast and Robust Method for Global Topological Functional Optimizationcited by 15
  19. 2019PI-Net: A Deep Learning Approach to Extract Topological Persistence Imagescited by 12
  20. 2017Assessing the Stylistic Properties of Neurally Generated Text in Authorship Attributioncited by 5
  21. 2020Generative Models are Unsupervised Predictors of Page Quality: A Colossal-Scale Studycited by 4
  22. 2020Loss Barcode: A Topological Measure of Escapability in Loss Landscapescited by 3

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 1; 2022: 12; 2023: 16; 2024: 9; 2025: 5; 2026: 6.

49 papers.

Sort:

By area:

  1. 2023Can LLM-Generated Misinformation Be Detected?cited by 173
  2. 2023Intrinsic Dimension Estimation for Robust Detection of AI-Generated Textscited by 111
  3. 2022Machine Generated Text: A Comprehensive Survey of Threat Models and Detection Methodscited by 75
  4. 2022Do Language Models Plagiarize?cited by 72
  5. 2023Stylometric Detection of AI-Generated Text in Twitter Timelinescited by 57
  6. 2022Attribution and Obfuscation of Neural Text Authorship: A Data Mining Perspectivecited by 40
  7. 2023GPT-who: An Information Density-based Machine-Generated Text Detectorcited by 38
  8. 2022Findings of the The RuATD Shared Task 2022 on Artificial Text Detection in Russiancited by 29
  9. 2023On the Detectability of ChatGPT Content: Benchmarking, Methodology, and Evaluation through the Lens of Academic Writingcited by 26
  10. 2023Does Human Collaboration Enhance the Accuracy of Identifying LLM-Generated Deepfake Texts?cited by 21
  11. 2023To ChatGPT, or not to ChatGPT: That is the question!cited by 20
  12. 2023AI-generated text boundary detection with RoFTcited by 20
37 more
  1. 2022Acceptability Judgements via Examining the Topology of Attention Mapscited by 19
  2. 2024A Survey of AI-generated Text Forensic Systems: Detection, Attribution, and Characterizationcited by 18
  3. 2025Hallucination Detection in LLMs with Topological Divergence on Attention Graphscited by 18
  4. 2023On the Trustworthiness Landscape of State-of-the-art Generative Models: A Survey and Outlookcited by 15
  5. 2022The Topological BERT: Transforming Attention into Topology for Natural Language Processingcited by 10
  6. 2023TOPFORMER: Topology-Aware Authorship Attribution of Deepfake Texts with Diverse Writing Stylescited by 10
  7. 2022Topological Data Analysis for Speech Processingcited by 9
  8. 2023An Empirical Study of AI Generated Text Detection Toolscited by 9
  9. 2023Uncertainty Estimation of Transformers' Predictions via Topological Analysis of the Attention Matricescited by 8
  10. 2022miCSE: Mutual Information Contrastive Learning for Low-shot Sentence Embeddingscited by 7
  11. 2022BERTops: Studying BERT Representations under a Topological Lenscited by 5
  12. 2022Dialogue Term Extraction using Transfer Learning and Topological Data Analysiscited by 5
  13. 2023Can BERT eat RuCoLA? Topological Data Analysis to Explaincited by 5
  14. 2024Deciphering Textual Authenticity: A Generalized Strategy through the Lens of Large Language Semantics for Detecting Human vs. Machine-Generated Textcited by 5
  15. 2022Artificial Text Detection with Multiple Training Strategiescited by 4
  16. 2025MGT-Prism: Enhancing Domain Generalization for Machine-Generated Text Detection via Spectral Alignmentcited by 4
  17. 2023Stacking the Odds: Transformer-Based Ensemble for AI-Generated Text Detectioncited by 3
  18. 2024Topological Perspectives on Optimal Multimodal Embedding Spacescited by 3
  19. 2024GigaCheck: Detecting LLM-generated Content via Object-Centric Span Localizationcited by 3
  20. 2021Abstraction, Reasoning and Deep Learning: A Study of the "Look and Say" Sequencecited by 2
  21. 2022Synthetic Text Detection: Systemic Literature Reviewcited by 2
  22. 2023DEMASQ: Unmasking the ChatGPT Wordsmithcited by 2
  23. 2023Detecting out-of-distribution text using topological features of transformer-based language modelscited by 2
  24. 2024Local Topology Measures of Contextual Language Model Latent Spaces With Applications to Dialogue Term Extractioncited by 2
  25. 2024Topological Data Analysis Applications in Natural Language Processing: A Surveycited by 2
  26. 2026Topological Signatures of Grokkingcited by 2
  27. 2024Beyond Simple Averaging: Improving NLP Ensemble Performance with Topological-Data-Analysis-Based Weightingcited by 1
  28. 2024Vulnerability Detection via Topological Analysis of Attention Mapscited by 1
  29. 2025Better Call Claude: Can LLMs Detect Changes of Writing Style?cited by 1
  30. 2025Two Birds with One Stone: Multi-Task Detection and Attribution of LLM-Generated Textcited by 1
  31. 2026The Density of Cross-Persistence Diagrams and Its Applicationscited by 1
  32. 2024Characterization of topological structures in different neural network architecturescited by 0
  33. 2025Detecting Narrative Shifts through Persistent Structures: A Topological Analysis of Media Discoursecited by 0
  34. 2026Optimizing Soft Prompt Tuning via Structural Evolutioncited by 0
  35. 2026SV-Detect: AI-generated Text Detection with Steering Vectorscited by 0
  36. 2026CodeTD: Topology of Attention Detects Hallucinations in Code LLMscited by 0
  37. 2026A Mechanistic Study of AI-Text Detection Neurons in Frozen BERT: Sparse Probing and Activation Patching on RAIDcited by 0