mlsanityData to 14 Oct 2026

Decoupling Content and Expression: Two-Dimensional Detection of AI-Generated Text

arXiv 2503.00258 ↗First posted 1 Mar 2025cs.CL cs.AI cs.CYcited by 6

Not on the map. Why?

Abstract (arXiv)

The wide usage of LLMs raises critical requirements on detecting AI participation in texts. Existing studies investigate these detections in scattered contexts, leaving a systematic and unified approach unexplored. In this paper, we present HART, a hierarchical framework of AI risk levels, each corresponding to a detection task. To address these tasks, we propose a novel 2D Detection Method, decoupling a text into content and language expression. Our findings show that content is resistant to surface-level changes, which can serve as a key feature for detection. Experiments demonstrate that 2D method significantly outperforms existing detectors, achieving an AUROC improvement from 0.705 to 0.849 for level-2 detection and from 0.807 to 0.886 for RAID. We release our data and code at https://github.com/baoguangsheng/truth-mirror.

Key facts. Cited by 6 papers in the map (data through 14 Oct 2026). Builds on 39 of its 66 references that are in the map. First posted on arXiv 1 Mar 2025.

Builds on

39 of 66 references.

References · 39

  1. 2019RoBERTa: A Robustly Optimized BERT Pretraining Approachcited by 13,372
  2. 2018Hierarchical Neural Story Generationcited by 1,193
  3. 2023DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvaturecited by 726
  4. 2020Knowledge Graphscited by 621
  5. 2019Release Strategies and the Social Impacts of Language Modelscited by 513
  6. 2019GLTR: Statistical Detection and Visualization of Generated Textcited by 496
  7. 2023How Close is ChatGPT to Human Experts? Comparison Corpus, Evaluation, and Detectioncited by 481
  8. 2023Paraphrasing evades detectors of AI-generated text, but retrieval is an effective defensecited by 429
  9. 2018Black-box Generation of Adversarial Text Sequences to Evade Deep Learning Classifierscited by 345
  10. 2023Fast-DetectGPT: Efficient Zero-Shot Detection of Machine-Generated Text via Conditional Probability Curvaturecited by 287
  11. 2024Spotting LLMs With Binoculars: Zero-Shot Detection of Machine-Generated Textcited by 263
  12. 2019Automatic Detection of Generated Text is Easiest when Humans are Fooledcited by 261
27 more
  1. 2022CoAuthor: Designing a Human-AI Collaborative Writing Dataset for Exploring Language Model Capabilitiescited by 249
  2. 2023DetectLLM: Leveraging Log Rank Information for Zero-Shot Detection of Machine-Generated Textcited by 210
  3. 2023RADAR: Robust AI-Text Detection via Adversarial Learningcited by 190
  4. 2023MAGE: Machine-generated Text Detection in the Wildcited by 132
  5. 2023DNA-GPT: Divergent N-Gram Analysis for Training-Free Detection of GPT-Generated Textcited by 129
  6. 2023M4: Multi-generator, Multi-domain, and Multi-lingual Black-Box Machine-Generated Text Detectioncited by 121
  7. 2024Mapping the Increasing Use of LLMs in Scientific Paperscited by 118
  8. 2020TweepFake: about Detecting Deepfake Tweetscited by 117
  9. 2023MGTBench: Benchmarking Machine-Generated Text Detectioncited by 93
  10. 2023Protecting Language Generation Models via Invisible Watermarkingcited by 90
  11. 2021Sparks: Inspiration for Science Writing using Language Modelscited by 80
  12. 2023A Survey on Detection of LLMs-Generated Contentcited by 71
  13. 2019Generating Sentences from Disentangled Syntactic and Semantic Spacescited by 59
  14. 2023ChatGPT and a New Academic Reality: Artificial Intelligence-Written Research Papers and the Ethics of the Large Language Models in Scholarly Publishingcited by 52
  15. 2024SoK: Watermarking for AI-Generated Contentcited by 51
  16. 2024LLM-as-a-Coauthor: Can Mixed Human-Written and Machine-Generated Text Be Detected?cited by 49
  17. 2021Disentangling Syntax and Semantics in the Brain with Deep Networkscited by 40
  18. 2023Fighting Fire with Fire: Can ChatGPT Detect AI-generated Text?cited by 39
  19. 2019A Multi-Task Approach for Disentangling Syntax and Semantics in Sentence Representationscited by 37
  20. 2024Weaver: Foundation Models for Creative Writingcited by 27
  21. 2024Stumbling Blocks: Stress Testing the Robustness of Machine-Generated Text Detectors Under Attackscited by 26
  22. 2024LLM-DetectAIve: a Tool for Fine-Grained Machine-Generated Text Detectioncited by 25
  23. 2024Detecting Subtle Differences between Human and Model Languages Using Spectrum of Relative Likelihoodcited by 24
  24. 2024StoryVerse: Towards Co-authoring Dynamic Plot with LLM-based Character Simulation via Narrative Planningcited by 20
  25. 2023How do humans perceive adversarial text? A reality check on the validity and naturalness of word-based adversarial attackscited by 18
  26. 2024Glimpse: Enabling White-Box Methods to Use Proprietary Models for Zero-Shot LLM-Generated Text Detectioncited by 14
  27. 2024LLMs as Writing Assistants: Exploring Perspectives on Sense of Ownership and Reasoningcited by 12

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 4; 2026: 2.

6 papers.

  1. 2025DETree: DEtecting Human-AI Collaborative Texts via Tree-Structured Hierarchical Representation Learningcited by 12
  2. 2025DetectAnyLLM: Towards Generalizable and Robust Detection of Machine-Generated Text Across Domains and Modelscited by 9
  3. 2025T-Detect: Tail-Aware Statistical Normalization for Robust Detection of Adversarial Machine-Generated Textcited by 1
  4. 2025AI-Generated Text is Non-Stationary: Detection via Temporal Tomographycited by 1
  5. 2026DWT-Fusion: A Signal-Based Framework for Training-Free LLM-Generated Text Detectioncited by 0
  6. 2026How Much Were You Told? Measuring External Information in Peer Reviewscited by 0