Elamur Research⌕
☰
home / papers / 2026 / WaveDetect: Robust Framework for Machine-Generated Text Detection via Wavelet Transform

WaveDetect: Robust Framework for Machine-Generated Text Detection via Wavelet Transform

Paper · 2026 · Zhichen Liu, Kaitong Qin, Linhan He et al.

Summary

Machine-generated text is becoming hard to separate from human writing. Detectors that rely on surface or semantic cues tend to break under three conditions: paraphrasing and other adversarial edits, shifts to new domains, and newer generator models. WaveDetect treats detection as a signal-processing problem. A surrogate causal language model scores each token, and the resulting probability sequence is treated as a signal. A differentiable Continuous Wavelet Transform with a Morlet mother wavelet turns that signal into a time–scale spectrogram, and a lightweight CNN classifies the spectrogram as human or LLM text.

Unlike Fourier-based approaches such as FourierGPT, wavelets keep information about where in the text a pattern occurs. The authors argue this matters because decoding strategies smooth machine probability signals, while human text shows irregular local spikes. The system is evaluated on RAID, EvoBench and Domain-Shift. It reports new state-of-the-art results, with notable robustness to attacks, unseen domains and newer LLMs. This suggests spectral analysis is a promising direction for building detectors that generalize.

Key points

  • Pipeline: surrogate LM token probabilities → CWT (implemented as a fixed-weight 1D convolution bank over scales) → spectral magnitude map → CNN classifier. Padded positions in the spectrum are zero-masked.
  • Training: two stages, first a warm-up with the surrogate model frozen, then joint fine-tuning of the surrogate and the CNN. Training uses a weighted cross-entropy loss to handle class imbalance.
  • Evaluation: RAID tests adversarial robustness, EvoBench tests model evolution, and Domain-Shift tests out-of-domain generalization, for example legal and medical text.
  • Interpretability:
    • Grad-CAM shows the classifier attends to small-scale (high-frequency) regions for human text and large-scale (low-frequency) regions for machine text.
    • t-SNE shows near-complete separation for GPT-4/ChatGPT text, although some paraphrased outputs overlap with the human cluster.
  • Limitations and code: training relied on RAID, whose generator models are relatively old, and the effects of wavelet choice and scale range remain unexplored. Code is available at github.com/KaitongQin/WaveDetect.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Institutions

🎓Southern University of Science and TechnologyShenzhen · China19 works · 1,337 citations

 Fields, topics and concepts

📚 Shelves 5The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 worksComputer Vision and Pattern Recognition · 2,869 works

Cite

Cite as Zhichen Liu, Kaitong Qin, Linhan He and Yang Xu. 2026. WaveDetect: Robust Framework for Machine-Generated Text Detection via Wavelet Transform. Findings of the Association for Computational Linguistics: ACL 2026.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.findings-acl.424

@inproceedings{liu2026wavedetect,
  title = {{WaveDetect: Robust Framework for Machine-Generated Text Detection via Wavelet Transform}},
  author = {Zhichen Liu, Kaitong Qin, Linhan He and Yang Xu},
  booktitle = {Findings of the Association for Computational Linguistics: ACL 2026},
  year = {2026},
  doi = {10.18653/v1/2026.findings-acl.424},
  url = {https://doi.org/10.18653/v1/2026.findings-acl.424}
}

References

49 references, 14 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Mervat Abassy, Kareem Elozeiri, Alexander Aziz, Minh Ngoc Ta, Raj Vardhan Tomar, Bimarsha Adhikari, Saad El Dine Ahmed, Yuxia Wang, Osama Mohammed Afzal, Zhuohan Xie, Jonibek Mansurov, Ekaterina Artemova, et al. 2024. LLM-DetectAIve: a tool for fine-grained machine-generated text detection. In Proce… find
  2. Micael Arman. 2020. Poems dataset (nlp). https://www.kaggle.com/datasets/ michaelarman/poemsdataset. Kaggle dataset. ↗
  3. David Bamman and Noah A. Smith. 2013. New alignment methods for discriminative book summarization. In Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing. find
  4. Guangsheng Bao, Yanbin Zhao, Zhiyang Teng, Linyi Yang, and Yue Zhang. 2024. Fast-detectgpt: Efficient zero-shot detection of machine-generated text via conditional probability curvature. Preprint, arXiv:2310.05130.
  5. Aaditya Bhat. 2023. https://doi.org/10.57967/hf/0326. Wikipedia-style introductions. — [doi:10.57967/hf/0326](https://doi.org/10.57967/hf/0326) ↗
  6. Gpt-wiki-intro. Dataset for Michał Bień, Michał Gilski, Martyna Maciejewska, Wojciech Taisner, Dawid Wiśniewski, and Agnieszka Ławrynowicz. 2020. Recipenlg: A cooking recipes dataset for semi-structured text generation. In Proceedings of the 13th International Conference on Natural Language Genera… find
  7. Matyáš Boháček, Michal Bravanský, Filip Trhlík, and Václav Moravec. 2022. Fine-grained czech news article dataset: An interdisciplinary approach to trustworthiness analysis. find
  8. Umar Butler. 2023. Open australian legal qa. find
  9. ICLR 2026 Program Chairs. 2025. find
  10. ICLR 2026 Response to LLM-Generated Papers and Reviews. find
  11. Xin Chen, Junchao Wu, Shu Yang, Runzhe Zhan, Zeyu Wu, Ziyang Luo, Di Wang, Min Yang, Lidia S. Chao, and Derek F. Wong. 2025a. Repreguard: Detecting llm-generated text by revealing hidden representation patterns. Preprint, arXiv:2508.13152. find
  12. Zhihui Chen, Kai He, Yucheng Huang, Yunxiao Zhu, and Mengling Feng. 2025b. Divscore: Zero-shot detection of llm-generated text in specialized domains. Preprint, arXiv:2506.06705. find
  13. Evan Crothers, Nathalie Japkowicz, and Herna Viktor. 2023. Machine-generated text: A comprehensive survey of threat models and detection methods. IEEE Access, PP:1–1. find
  14. DeepSeek-AI, Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, Xiaokang Zhang, Xingkai Yu, Yu Wu, Z. F. Wu, Zhibin Gou, Zhihong Shao, Zhuoshu Li, Ziyi Gao, Aixin Liu, et al. 2025a. Deepseek-r1: Incentivizing reasoning capability in…
  15. DeepSeek-AI, Aixin Liu, Bei Feng, Bing Xue, Bingxuan Wang, Bochao Wu, Chengda Lu, Chenggang Zhao, Chengqi Deng, Chenyu Zhang, Chong Ruan, Damai Dai, Daya Guo, Dejian Yang, Deli Chen, Dongjie Ji, Erhang Li, Fangyun Lin, Fucong Dai, Fuli Luo, et al. 2025b. Deepseek-v3 technical report. Preprint, arXiv… find
  16. Liam Dugan, Alyssa Hwang, Filip Trhlík, Andrew Zhu, Josh Magnus Ludan, Hainiu Xu, Daphne Ippolito, and Chris Callison-Burch. 2024. RAID: A shared benchmark for robust evaluation of machine-generated text detectors.
  17. Angela Fan, Mike Lewis, and Yann Dauphin. 2018. Hierarchical neural story generation. In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 889–898, Melbourne, Australia. Association for Computational Linguistics.
  18. Sebastian Gehrmann, Hendrik Strobelt, and Alexander Rush. 2019. GLTR: Statistical detection and visualization of generated text. In Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics: System Demonstrations, pages 111–116, Florence, Italy. Association for Computat…
  19. Derek Greene and Pádraig Cunningham. 2006. Practical solutions to the problem of diagonal dominance in kernel document clustering. In Proceedings of the 23rd International Conference on Machine Learning (ICML), pages 377–384. ACM. find
  20. Alexander Grossmann and Jean Morlet. 1984. Decomposition of hardy functions into square integrable wavelets of constant shape. Siam Journal on Mathematical Analysis, 15:723–736. find
  21. Xun Guo, Shan Zhang, Yongxin He, Ting Zhang, Wanquan Feng, Haibin Huang, and Chongyang Ma. 2024. Detective: Detecting ai-generated text via multi-level contrastive learning. In Advances in Neural Information Processing Systems, volume 37, pages 88320– 88347. Curran Associates, Inc. find
  22. Abhimanyu Hans, Avi Schwarzschild, Valeriia Cherepanova, Hamid Kazemi, Aniruddha Saha, Micah Goldblum, Jonas Geiping, and Tom Goldstein. 2024. Spotting llms with binoculars: Zero-shot detection of machine-generated text. Preprint, arXiv:2401.12070. find
  23. Xiaomeng Hu, Pin-Yu Chen, and Tsung-Yi Ho. 2023. Radar: Robust ai-text detection via adversarial learning. Preprint, arXiv:2307.03838. find
  24. Albert Q. Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lucile Saulnier, Lélio Renard Lavaud, Marie-Anne Lachaux, Pierre Stock, Teven Le Scao, Thibaut Lavril, Thomas Wang, Timothée Lacroix,… find
  25. Qiao Jin, Bhuwan Dhingra, Zhengping Liu, William W. Cohen, and Xinghua Lu. 2019. Pubmedqa: A dataset for biomedical research question answering. Preprint, arXiv:1909.06146.
  26. Alistair E. W. Johnson, Lucas Bulgarelli, Lu Shen, Alvin Gayles, Ayad Shammout, Steven Horng, Tom J. Pollard, Sicheng Hao, Benjamin Moody, Brian Gow, Li-wei H. Lehman, Leo A. Celi, and Roger G. Mark. 2023. Mimic-iv, a freely accessible electronic health record dataset. Scientific Data, 10(1):1.
  27. Zhichen Liu, Yongyuan Li, Yang Xu, Yu Wang, Yingfang Yuan, and Zuhao Yang. 2025. Evaluating text generation quality using spectral distances of surprisal. In Findings of the Association for Computational Linguistics: EMNLP 2025, pages 2444–2463, Suzhou, China. Association for Computational Linguisti… find
  28. Andrew L. Maas, Raymond E. Daly, Peter T. Pham, Dan Huang, Andrew Y. Ng, and Christopher Potts. 2011. Learning word vectors for sentiment analysis. In Proceedings of the 49th Annual Meeting of the Association for Computational Linguistics: Human Language Technologies, pages 142–150, Portland, Oregon…
  29. S.G. Mallat. 1989. A theory for multiresolution signal decomposition: the wavelet representation. IEEE Transactions on Pattern Analysis and Machine Intelligence, 11(7):674–693. find
  30. Eric Mitchell, Yoonho Lee, Alexander Khazatsky, Christopher D. Manning, and Chelsea Finn. 2023. Detectgpt: Zero-shot machine-generated text detection using probability curvature. Preprint, arXiv:2301.11305.
  31. MosaicML. 2023. Mpt-30b: Raising the bar for open-source foundation models. find
  32. Yasmin Moslem. 2023. Law-stackexchange dataset. find
  33. Shashi Narayan, Shay B. Cohen, and Mirella Lapata. 2018. Don’t give me the details, just the summary! topic-aware convolutional neural networks for extreme summarization. ArXiv, abs/1808.08745.
  34. Long Ouyang, Jeff Wu, Xu Jiang, Diogo Almeida, Carroll L. Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, John Schulman, Jacob Hilton, Fraser Kelton, Luke Miller, Maddie Simens, Amanda Askell, Peter Welinder, Paul Christiano, Jan Leike, and Ryan Lowe. 2022. Train…
  35. Sayak Paul and Soumik Rakshit. 2021. arxiv paper abstracts. https://www.kaggle.com/datasets/ spsayakpaul/arxiv-paper-abstracts. Kaggle dataset. ↗
  36. Alec Radford, Jeffrey Wu, Rewon Child, David Luan, Dario Amodei, Ilya Sutskever, et al. 2019. Language models are unsupervised multitask learners. OpenAI blog, 1(8):9. find
  37. Veselin Raychev, Pavol Bielik, and Martin Vechev. 2016. Probabilistic model for code with decision trees. SIG-PLAN Notices, 51(10):731–747. find
  38. Dietmar Schabus, Marcin Skowron, and Martin Trapp. 2017. One million posts: A data set of german online discussions. In Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Retrieval, pages 1241–1244. Association for Computing Machinery. find
  39. Ramprasaath R. Selvaraju, Michael Cogswell, Abhishek Das, Ramakrishna Vedantam, Devi Parikh, and Dhruv Batra. 2019. Grad-cam: Visual explanations from deep networks via gradient-based localization. International Journal of Computer Vision, 128(2):336–359. find
  40. Irene Solaiman, Miles Brundage, Jack Clark, Amanda Askell, Ariel Herbert-Voss, Jeff Wu, Alec Radford, Gretchen Krueger, Jong Wook Kim, Sarah Kreps, Miles McCain, Alex Newhouse, Jason Blazakis, Kris McGuffie, and Jasmine Wang. 2019. Release strategies and the social impacts of language models. Prepri…
  41. Xiao Yu, Yi Yu, Dongrui Liu, Kejiang Chen, Weiming Zhang, Nenghai Yu, and Jing Shao. 2025. EvoBench: Towards real-world LLM-generated text detection benchmarking for evolving large language models. In Findings of the Association for Computational Linguistics: ACL 2025, pages 14605–14620, Vienna, Aus… find
  42. Qwen Team. 2024. Qwen2.5: A party of foundation models. find
  43. Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, Dan Bikel, Lukas Blecher, Cristian Canton Ferrer, Moya Chen, Guillem Cucurull, David Esiobu, Jude Fernandes, Jeremy Fu, et al. 2023. Llama 2: Op…
  44. Michael Völske, Martin Potthast, Shahbaz Syed, and Benno Stein. 2017. Tl;dr: Mining reddit to learn automatic summarization. In Proceedings of the Workshop on New Frontiers in Summarization, pages 59–63, Copenhagen, Denmark. Association for Computational Linguistics.
  45. Yichen Wang, Shangbin Feng, Abe Bohan Hou, Xiao Pu, Chao Shen, Xiaoming Liu, Yulia Tsvetkov, and Tianxing He. 2024. Stumbling blocks: Stress testing the robustness of machine-generated text detectors under attacks. Preprint, arXiv:2402.11638. find
  46. Alva West, Yixuan Weng, Minjun Zhu, Luodan Zhang, Zhen Lin, Guangsheng Bao, and Yue Zhang. 2025. Ai-generated text is non-stationary: Detection via temporal tomography. Preprint, arXiv:2508.01754. — [arXiv:2508.01754](https://arxiv.org/abs/2508.01754) find
  47. Yang Xu, Yu Wang, Hao An, Zhichen Liu, and Yongyuan Li. 2024. Detecting subtle differences between human and model languages using spectrum of relative likelihood. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 10108–10121, Miami, Florida, USA. Assoc… find
  48. Xianjun Yang, Liangming Pan, Xuandong Zhao, Haifeng Chen, Linda Ruth Petzold, William Yang Wang, and Wei Cheng. 2024. A survey on detection of LLMs-generated content. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 9786–9805, Miami, Florida, USA. Association for Compu… find
  49. Zuhao Yang, Yingfang Yuan, Yang Xu, Shuo Zhan, Huajun Bai, and Kefan Chen. 2023. Face: Evaluating natural language generation with fourier analysis of cross-entropy. Preprint, arXiv:2305.10307. — [arXiv:2305.10307](https://arxiv.org/abs/2305.10307) find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/liu-2026-wavedetect-robust-framework-machine-generated/