Authorship Attribution for Neural Text Generation
Open accessAbstract
In recent years, the task of generating realistic short and long texts have made tremendous advancements. In particular, several recently proposed neural network-based language models have demonstrated their astonishing capabilities to generate texts that are challenging to distinguish from human-written texts with the naked eye. Despite many benefits and utilities of such neural methods, in some applications, being able to tell the “author” of a text in question becomes critically important. In this work, in the context of this Turing Test, we investigate the so-called authorship attribution problem in three versions: (1) given two texts T1 and T2, are both generated by the same method or not? (2) is the given text T written by a human or machine? (3) given a text T and k candidate neural methods, can we single out the method (among k alternatives) that generated T? Against one humanwritten and eight machine-generated texts (i.e., CTRL, GPT, GPT2, GROVER, XLM, XLNET, PPLM, FAIR), we empirically experiment with the performance of various models in three problems. By and large, we find that most generators still generate texts significantly different from human-written ones, thereby making three problems easier to solve. However, the qualities of texts generated by GPT2, GROVER, and FAIR are better, often confusing machine classifiers in solving three problems. All codes and datasets of our experiments are available at: https://bit.ly/ 302zWdz
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.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
LLM-generated content often exhibits inherent biases and can be misused for malicious purposes.
“Additionally, LLM-generated content often exhibits inherent biases [20] and can be misused for malicious purposes [37,53], further complicating efforts to maintain information integrity.”
From Beyond Binary: Towards Fine-Grained LLM-Generated Text Detection via Role Recognition and Involvement Measurement (2025) · cites this paper as [53]
This source is cited for the malicious-use clause; the bias clause is cited to [20]. The available passages contain only title and bibliographic metadata, with no substantive text addressing misuse. Searches for malicious purposes, misinformation, and propaganda found no relevant passage, so the attribution cannot be verified from our copy. (LLM, conf=0.72)
Experimental — accuracy not yet measuredNeural network-based detectors face limitations such as overfitting to training data.
“These detectors also face limitations such as overfitting to training data [Uchendu et al., 2020].”
From Authorship Attribution in the Era of LLMs: Problems, Methodologies, and Challenges (2025) · cites this paper as Uchendu et al., 2020
The supplied passages contain bibliographic metadata, not a discussion of detector limitations. Searches for overfitting, generalization, and training-data limitations returned only the title and author listing (no section or page), so the available copy does not establish this claim. (LLM, conf=0.72)
Experimental — accuracy not yet measuredEncoder-based fine-tuning approaches lack robustness, as they tend to overfit to their training data or the source model's training distribution, resulting in a decline in performance when faced with cross-domain or unseen data.
“Nevertheless, like their counterparts, these encoder-based fine-tuning approaches lack robustness (Bakhtin et al. 2019; Uchendu et al. 2020; Antoun et al. 2023a; Li et al. 2023c), as they tend to overfit to their training data or the source model's training distribution, resulting in a decline in performance when faced with cross-domain or unseen data.”
From A Survey on LLM-Generated Text Detection: Necessity, Methods, and Future Directions (2025) · cites this paper as Uchendu et al. 2020
The indexed passages contain the paper’s title and bibliographic metadata, but no substantive text addressing encoder-based fine-tuning. Searches for overfitting, cross-domain generalization, and performance on unseen data returned no relevant passage, so the attributed claim cannot be checked against our copy. (LLM, conf=0.72)
Experimental — accuracy not yet measured
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
RoBERTa: A Robustly Optimized BERT Pretraining Approach
Liu, Y., Ott, M., Goyal, N., Du, J., Joshi, M., Chen, D., Levy, O., Lewis, M., Zettlemoyer, L., & Stoyanov, V. (2019) · arXiv · doi:10.48550/arxiv.1907.11692 · Cited by 16,515 · see citing works
Authorship Attribution with Topic Models
Seroussi, Y., Zukerman, I., & Bohnert, F. (2014) · Computational Linguistics · doi:10.1162/coli_a_00173 · Cited by 89 · see citing works
Sharma, A., Nandan, A., & Ralhan, R. (2018) · arXiv · doi:10.48550/arxiv.1812.10281 · Cited by 7 · see citing works
Flagged first, then by citations
Cited by
Sorted by citations
All That’s ‘Human’ Is Not Gold: Evaluating Human Evaluation of Generated Text
Clark, E., August, T., Serrano, S., Haduong, N., Gururangan, S., & Smith, N. A. (2021) · Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing (Volume 1: Long Papers) · doi:10.18653/v1/2021.acl-long.565 · Cited by 290 · see citing works
All That's 'Human' Is Not Gold: Evaluating Human Evaluation of Generated Text
Clark, E., August, T., Serrano, S., Haduong, N., Gururangan, S., & Smith, N. A. (2021) · arXiv · doi:10.48550/arxiv.2107.00061 · Cited by 290
A Survey on LLM-Generated Text Detection: Necessity, Methods, and Future Directions
Wu, J., Yang, S., Zhan, R., Yuan, Y., Chao, L. S., & Wong, D. F. (2025) · Computational Linguistics · doi:10.1162/coli_a_00549 · Cited by 212 · see citing works
DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvature
Mitchell, E., Lee, Y., Khazatsky, A., Manning, C. D., & Finn, C. (2023) · arXiv · doi:10.48550/arxiv.2301.11305 · Cited by 151 · see citing works
MAGE: Machine-generated Text Detection in the Wild
Li, Y., Li, Q., Cui, L., Bi, W., Wang, Z., Wang, L., Yang, L., Shi, S., & Zhang, Y. (2024) · Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · doi:10.18653/v1/2024.acl-long.3 · Cited by 79 · see citing works
Quantifying large language model usage in scientific papers
Liang, W., Zhang, Y., Wu, Z., Lepp, H., Ji, W., Zhao, X., Cao, H., Liu, S., He, S., Huang, Z., et al. (2025) · Nature Human Behaviour · doi:10.1038/s41562-025-02273-8 · Cited by 80 · see citing works
Liang, W., Izzo, Z., Zhang, Y., Lepp, H., Cao, H., Zhao, X., Chen, L., Ye, H., Liu, S., Huang, Z., et al. (2024) · arXiv · doi:10.48550/arxiv.2403.07183 · Cited by 73 · see citing works
“Earlier studies have used classifiers to detect synthetic text in peer review corpora (Bhagat & Hovy, 2013), media outlets (Zellers et al., 2019), and other contexts (Bakhtin et al., 2019; Uchendu et al., 2020).”
Ghostbuster: Detecting Text Ghostwritten by Large Language Models
Verma, V., Fleisig, E., Tomlin, N., & Klein, D. (2024) · Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers) · doi:10.18653/v1/2024.naacl-long.95 · Cited by 56 · see citing works
SeqXGPT: Sentence-Level AI-Generated Text Detection
Wang, P., Li, L., Ren, K., Jiang, B., Zhang, D., & Qiu, X. (2023) · Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing · doi:10.18653/v1/2023.emnlp-main.73 · Cited by 53 · see citing works
ConDA: Contrastive Domain Adaptation for AI-generated Text Detection
Bhattacharjee, A., Kumarage, T., Moraffah, R., & Liu, H. (2023) · Proceedings of the 13th International Joint Conference on Natural Language Processing and the 3rd Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (Volume 1: Long Papers) · doi:10.18653/v1/2023.ijcnlp-main.40 · Cited by 43 · see citing works
Cross-domain detection of GPT-2-generated technical text
Rodriguez, J., Hay, T., Gros, D., Shamsi, Z., & Srinivasan, R. (2022) · Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies · doi:10.18653/v1/2022.naacl-main.88 · Cited by 40 · see citing works
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
“These detectors also face limitations such as overfitting to training data (Uchendu et al., 2020).”
Self-citationM4GT-Bench: Evaluation Benchmark for Black-Box Machine-Generated Text Detection
Wang, Y., Mansurov, J., Ivanov, P., Su, J., Shelmanov, A., Tsvigun, A., Mohammed Afzal, O., Mahmoud, T., Puccetti, G., Arnold, T., et al. (2024) · Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · doi:10.18653/v1/2024.acl-long.218 · Cited by 25 · see citing works
Malla: Demystifying Real-world Large Language Model Integrated Malicious Services
Lin, Z., Cui, J., Liao, X., & Wang, X. (2024) · arXiv · doi:10.48550/arxiv.2401.03315 · Cited by 13 · see citing works
“To uncover the backend LLMs employed by three Malla services (i.e., DarkGPT, EscapeGPT, and FreedomGPT), we explore techniques to address the LLM authorship attribution problem [88,123], i.e., given a set of responses T and k candidate LLMs, what is the LLM (among k alternatives) that generated T ?”
MethodMAGE: Machine-generated Text Detection in the Wild
Li, Y., Li, Q., Cui, L., Bi, W., Wang, Z., Wang, L., Yang, L., Shi, S., & Zhang, Y. (2023) · arXiv · doi:10.48550/arxiv.2305.13242 · Cited by 10 · see citing works
Cheng, Z., Zhou, L., Jiang, F., Wang, B., & Li, H. (2025) · Proceedings of the ACM on Web Conference 2025 · doi:10.1145/3696410.3714770 · Cited by 3 · see citing works
Towards Automatic Boundary Detection for Human-AI Collaborative Hybrid Essay in Education
Zeng, Z., Sha, L., Li, Y., Yang, K., Gašević, D., & Chen, G. (2023) · arXiv · doi:10.48550/arxiv.2307.12267 · Cited by 25 · see citing works
- Authors
- Uchendu, A., Le, T., Shu, K., & Lee, D.
- Venue
- Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)
- Published
- 2020
- Pages
- 8384-8395
- OpenAlex
- W3101891351
- Cited by
- 120
- References
- 3
- Citations by year
- 2026: 12 · 2025: 31 · 2024: 29 · 2023: 26 · 2022: 11 · 2021: 9 · 2020: 2
- FWCI
- 4.58
- Type
- conference paper
- Language
- en
- Open access
- hybrid
- doi.org
- Licence
- cc-by
- Locations
- Conference on Empirical Methods in Natural Language Processing (EMNLP) (publishedVersion, cc-by)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Authorship Attribution and Profiling
- Physical Sciences › Computer Science › Artificial Intelligence › Topic Modeling
- Physical Sciences › Computer Science › Artificial Intelligence › Natural Language Processing Techniques
- Keywords
- authorship attribution, neural text generation, AI-generated text detection, Turing test, language models, GPT-2, GROVER, XLNet, XLM
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
