Peer Review #1 of "Feature-based detection of automated language models: tackling GPT-2, GPT-3 and Grover (v0.1)"
Open accessAbstract
The recent improvements of language models have drawn much attention to potential cases of use and abuse of automatically generated text.Great effort is put into the development of methods to detect machine generations among human-written text in order to avoid scenarios in which the large-scale generation of text with minimal cost and effort undermines the trust in human interaction and factual information online.While most of the current approaches rely on the availability of expensive language models, we propose a simple feature-based classifier for the detection problem, using carefully crafted features that attempt to model intrinsic differences between human and machine text.Our research contributes to the field in producing a detection method that achieves performance competitive with far more expensive methods, offering an accessible "first line-of-defense" against the abuse of language models.Furthermore, our experiments show that different sampling methods lead to different types of flaws in generated text.
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Cites
Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2017) · arXiv · doi:10.48550/arxiv.1706.03762 · Cited by 120,404 · see citing works
Distributed Representations of Words and Phrases and their Compositionality
Mikolov, T., Sutskever, I., Chen, K., Corrado, G., & Dean, J. (2013) · arXiv · doi:10.48550/arxiv.1310.4546 · Cited by 24,533 · see citing works
Language Models are Few-Shot Learners
Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., et al. (2020) · arXiv · doi:10.48550/arxiv.2005.14165 · Cited by 23,372 · see citing works
An introduction to ROC analysis
Fawcett, T. (2006) · Pattern Recognition Letters · doi:10.1016/j.patrec.2005.10.010 · Cited by 21,879 · see citing works
Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
Raffel, C., Shazeer, N., Roberts, A., Lee, K., Narang, S., Matena, M., Zhou, Y., Li, W., & Liu, P. J. (2019) · arXiv · doi:10.48550/arxiv.1910.10683 · Cited by 10,929 · see citing works
Text categorization with Support Vector Machines: Learning with many relevant features
Joachims, T. (1998) · Lecture notes in computer science · doi:10.1007/bfb0026683 · Cited by 8,134 · see citing works
Machine learning in automated text categorization
Sebastiani, F. (2002) · ACM Computing Surveys · doi:10.1145/505282.505283 · Cited by 7,927 · see citing works
SciBERT: A Pretrained Language Model for Scientific Text
Beltagy, I., Lo, K., & Cohan, A. (2019) · Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP) · doi:10.18653/v1/d19-1371 · Cited by 3,218 · see citing works
A re-examination of text categorization methods
Yang, Y., & Liu, X. (1999) · International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR) · doi:10.1145/312624.312647 · Cited by 2,670 · see citing works
The Curious Case of Neural Text Degeneration
Holtzman, A., Buys, J., Du, L., Forbes, M., & Choi, Y. (2019) · arXiv · doi:10.48550/arxiv.1904.09751 · Cited by 1,304 · see citing works
Japanese and Korean voice search
Schuster, M., & Nakajima, K. (2012) · 2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp.2012.6289079 · Cited by 862 · see citing works
Automatic Detection of Fake NewsPreprint
Pérez‐Rosas, V., Kleinberg, B., Lefevre, A., & Mihalcea, R. F. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1708.07104 · Cited by 709 · see citing works
Modeling Local Coherence: An Entity-Based Approach
Barzilay, R., & Lapata, M. (2008) · Computational Linguistics · doi:10.1162/coli.2008.34.1.1 · Cited by 673 · see citing works
Automatically Categorizing Written Texts by Author Gender
Koppel, M. (2002) · Literary and Linguistic Computing · doi:10.1093/llc/17.4.401 · Cited by 623 · see citing works
Fake News or Truth? Using Satirical Cues to Detect Potentially Misleading News
Rubin, V. L., Conroy, N. J., Chen, Y., & Cornwell, S. (2016) · doi:10.18653/v1/w16-0802 · Cited by 598 · see citing works
Defending Against Neural Fake News
Zellers, R., Holtzman, A., Rashkin, H., Bisk, Y., Farhadi, A., Roesner, F., & Choi, Y. (2019) · arXiv · doi:10.48550/arxiv.1905.12616 · Cited by 569 · see citing works
GLTR: Statistical Detection and Visualization of Generated Text
Gehrmann, S., Strobelt, H., & Rush, A. (2019) · Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics: System Demonstrations · doi:10.18653/v1/p19-3019 · Cited by 470 · see citing works
Empath: Understanding Topic Signals in Large-Scale Text
Fast, E., Chen, B., & Bernstein, M. S. (2016) · Proceedings of the 2016 CHI Conference on Human Factors in Computing Systems · doi:10.1145/2858036.2858535 · Cited by 413 · see citing works
Bisk, Y., Holtzman, A., Thomason, J., Andreas, J., Bengio, Y., Chai, J. Y., Lapata, M., Lazaridou, A., May, J., Nisnevich, A., et al. (2020) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/2020.emnlp-main.703 · Cited by 325 · see citing works
Release Strategies and the Social Impacts of Language Models
Solaiman, I., Brundage, M., Clark, J., Askell, A., Herbert-Voss, A., Wu, J., Radford, A., Krueger, G., Kim, J. W., Kreps, S., et al. (2019) · arXiv · doi:10.48550/arxiv.1908.09203 · Cited by 292 · see citing works
Automatic Detection of Generated Text is Easiest when Humans are Fooled
Ippolito, D., Duckworth, D., Callison-Burch, C., & Eck, D. (2019) · arXiv · doi:10.48550/arxiv.1911.00650 · Cited by 264
Automatic Detection of Generated Text is Easiest when Humans are Fooled
Ippolito, D., Duckworth, D., Callison-Burch, C., & Eck, D. (2020) · Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics · doi:10.18653/v1/2020.acl-main.164 · Cited by 264 · see citing works
Predicting Factuality of Reporting and Bias of News Media Sources
Baly, R., Karadzhov, G., Alexandrov, D., Glass, J. R., & Nakov, P. (2018) · Conference on Empirical Methods in Natural Language Processing (EMNLP) · doi:10.18653/v1/d18-1389 · Cited by 260 · see citing works
Automated Fact Checking: Task formulations, methods and future directionsPreprint
Thorne, J. P., & Vlachos, A. (2018) · arXiv (Cornell University) · doi:10.17863/cam.78745 · Cited by 245 · see citing works
A Comparison of Features for Automatic Readability Assessment
Feng, L., Jansche, M., Huenerfauth, M., & Elhadad, N. (2010) · Cited by 222 · see citing works
Text readability and intuitive simplification: A comparison of readability formulas
Crossley, S. A., Allen, D. B., & McNamara, D. S. (2011) · Reading in a Foreign Language · doi:10.64152/10125/66657 · Cited by 178 · see citing works
Unifying Human and Statistical Evaluation for Natural Language Generation
Hashimoto, T. B., Zhang, H., & Liang, P. (2019) · Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) · doi:10.18653/v1/n19-1169 · Cited by 161 · see citing works
A new approach to bot detection: Striking the balance between precision and recall
Morstatter, F., Wu, L., Nazer, T. H., Carley, K. M., & Liu, H. (2016) · IEEE/ACM International Conference on Advances in Social Networks Analysis and Mining (ASONAM) · doi:10.1109/asonam.2016.7752287 · Cited by 158 · see citing works
Do Massively Pretrained Language Models Make Better Storytellers?
See, A., Pappu, A., Saxena, R., Yerukola, A., & Manning, C. D. (2019) · Proceedings of the 23rd Conference on Computational Natural Language Learning (CoNLL) · doi:10.18653/v1/k19-1079 · Cited by 139 · see citing works
Style-based Text Categorization: What Newspaper Am I Reading?
Avneri, G. (1998) · Cited by 85 · see citing works
Real or Fake? Learning to Discriminate Machine from Human Generated Text
Bakhtin, A., Gross, S., Ott, M., Deng, Y., Ranzato, M., & Szlam, A. (2019) · arXiv · doi:10.48550/arxiv.1906.03351 · Cited by 77 · see citing works
Identifying Real or Fake Articles: Towards better Language Modeling
Badaskar, S., Agarwal, S., & Arora, S. (2008) · Cited by 31 · see citing works
Evaluating Machines by their Real-World Language UsePreprint
Zellers, R., Holtzman, A., Clark, E., Qin, L., Farhadi, A., & Choi, Y. (2020) · arXiv (Cornell University) · Cited by 15 · see citing works
Are We Safe Yet? The Limitations of Distributional Features for Fake News Detection.Preprint
Schuster, T., Schuster, R., Shah, D., & Barzilay, R. (2019) · arXiv (Cornell University) · Cited by 13 · see citing works
TLDR: Token Loss Dynamic Reweighting for Reducing Repetitive Utterance GenerationPreprint
Jiang, S., Wolf, T., Monz, C., & de Rijke, M. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2003.11963 · Cited by 13 · see citing works
Fully Automatic Journalism: We Need to Talk About Nonfake News Generation
Belz, A. (2019) · doi:10.36370/tto.2019.29 · Cited by 10 · see citing works
The Offense-Defense Balance of Scientific Knowledge: Does Publishing AI Research Reduce Misuse?Preprint
Shevlane, T., & Dafoe, A. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.2001.00463 · Cited by 9 · see citing works
EEGtoText: Learning to Write Medical Reports from EEG Recordings.
Biswal, S., Xiao, C., Westover, M. B., & Sun, J. (2019) · Machine Learning for Healthcare Conference · Cited by 9 · see citing works
Learning Within-Sentence Semantic Coherence
Eneva, E., Hoberman, R., & Lita, L. V. (2001) · Empirical Methods in Natural Language Processing · Cited by 8 · see citing works
TEASPN: Framework and Protocol for Integrated Writing Assistance Environments
Hagiwara, M., Ito, T., Kuribayashi, T., Suzuki, J., & Inui, K. (2019) · Conference on Empirical Methods in Natural Language Processing: System Demonstrations (EMNLP) · doi:10.18653/v1/d19-3039 · Cited by 7 · see citing works
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Devlin, J., Chang, M.-W., Lee, K., & Toutanova, K. (2018) · arXiv · doi:10.48550/arxiv.1810.04805
Flagged first, then by citations
Research works printed but not linked: 6
- Budzianowski, P. and Vulić, I. (2019). Hello, it's gpt-2-how can i help you? towards the use of pretrained language models for task-oriented dialogue systems. In Proceedings of the 3rd Workshop on Neural Generation and Translation, pages 15-22.
- Ferrara, E., Varol, O., Davis, C., Menczer, F., and Flammini, A. (2016). The rise of social bots. Communications of the ACM, 59(7):96-104.
- Lewandowsky, S., Ecker, U. K., Seifert, C. M., Schwarz, N., and Cook, J. (2012). Misinformation and its correction: Continued influence and successful debiasing. Psychological science in the public interest, 13(3):106-131.
- Radford, A., Wu, J., Child, R., Luan, D., Amodei, D., and Sutskever, I. (2019). Language models are unsupervised multitask learners. OpenAI Blog, 1(8):9.
- Sun, Z., Schuster, R., and Shmatikov, V. (2020). De-anonymizing text by fingerprinting language generation. arXiv preprint arXiv:2006.09615. · arXiv
- Zhang, T. and Oles, F. J. (2001). Text categorization based on regularized linear classification methods. Information retrieval, 4(1):5-31. 15/17 PeerJ Comput. Sci. reviewing PDF | (CS-2020:10:54411:1:3:NEW 18 Feb 2021) Manuscript to be reviewed Computer Science
Cited by
Sorted by citations
Authorship Attribution in the Era of LLMs: Problems, Methodologies, and Challenges
Huang, B., Chen, C., & Shu, K. (2025) · ACM SIGKDD Explorations Newsletter · doi:10.1145/3715073.3715076 · Cited by 42 · see citing works
“Similar to the problem of human authorship attribution, linguistic features such as phrasal verbs, coreference, part-of-speech (POS) tags, and named entity (NE) tags are also useful in distinguishing LLM-generated text (Fröhling and Zubiaga, 2021, Nguyen-Son et al., 2017, See et al., 2019).”
Understanding the effects of human-written paraphrases in LLM-generated text detection
Lau, H. T., & Zubiaga, A. (2025) · Natural Language Processing Journal · doi:10.1016/j.nlp.2025.100151 · Cited by 13 · see citing works
- Venue
- PeerJ Computer Science
- Publisher
- PeerJ
- Published
- 2021-04-06
- ISSN
- 2376-5992 (ISSN-L)
- OpenAlex
- W3156966624
- Cited by
- 0
- References
- 41
- Type
- peer review
- Language
- en
- Open access
- gold
- doi.org
- Licence
- cc-by
- Locations
- PeerJ Computer Science (publishedVersion, cc-by)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Authorship Attribution and Profiling
- Physical Sciences › Computer Science › Artificial Intelligence › Natural Language Processing Techniques
- Physical Sciences › Computer Science › Information Systems › Spam and Phishing Detection
- Keywords
- AI-generated text detection, feature-based classifiers, GPT-2, GPT-3, GROVER, sampling methods
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
