# Few-Shot Detection of Machine-Generated Text using Style Representations

arXiv: https://arxiv.org/abs/2401.06712 · HTML: https://mlsanity.com/paper/2401.06712/

## Key facts

- Cited by 40 papers in the map (data through 14 Oct 2026).
- Builds on 25 of its 42 references that are in the map.
- First posted on arXiv 12 Jan 2024.

## Builds on (most cited first, 25 of 25)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks](https://mlsanity.com/paper/1908.10084/) (2019)
- [On Calibration of Modern Neural Networks](https://mlsanity.com/paper/1706.04599/) (2017)
- [Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks](https://mlsanity.com/paper/1703.03400/) (2017)
- [OPT: Open Pre-trained Transformer Language Models](https://mlsanity.com/paper/2205.01068/) (2022)
- [Prototypical Networks for Few-shot Learning](https://mlsanity.com/paper/1703.05175/) (2017)
- [UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction](https://mlsanity.com/paper/1802.03426/) (2018)
- [DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvature](https://mlsanity.com/paper/2301.11305/) (2023)
- [Defending Against Neural Fake News](https://mlsanity.com/paper/1905.12616/) (2019)
- [Release Strategies and the Social Impacts of Language Models](https://mlsanity.com/paper/1908.09203/) (2019)
- [GLTR: Statistical Detection and Visualization of Generated Text](https://mlsanity.com/paper/1906.04043/) (2019)
- [Underspecification Presents Challenges for Credibility in Modern Machine Learning](https://mlsanity.com/paper/2011.03395/) (2020)
- [Do Deep Generative Models Know What They Don't Know?](https://mlsanity.com/paper/1810.09136/) (2018)
- [Paraphrasing evades detectors of AI-generated text, but retrieval is an effective defense](https://mlsanity.com/paper/2303.13408/) (2023)
- [Can AI-Generated Text be Reliably Detected?](https://mlsanity.com/paper/2303.11156/) (2023)
- [The Pushshift Reddit Dataset](https://mlsanity.com/paper/2001.08435/) (2020)
- [Automatic Detection of Generated Text is Easiest when Humans are Fooled](https://mlsanity.com/paper/1911.00650/) (2019)
- [Automatic Detection of Machine Generated Text: A Critical Survey](https://mlsanity.com/paper/2011.01314/) (2020)
- [M4: Multi-generator, Multi-domain, and Multi-lingual Black-Box Machine-Generated Text Detection](https://mlsanity.com/paper/2305.14902/) (2023)
- [Same Author or Just Same Topic? Towards Content-Independent Style Representations](https://mlsanity.com/paper/2204.04907/) (2022)
- [CATER: Intellectual Property Protection on Text Generation APIs via Conditional Watermarks](https://mlsanity.com/paper/2209.08773/) (2022)
- [RoFT: A Tool for Evaluating Human Detection of Machine-Generated Text](https://mlsanity.com/paper/2010.03070/) (2020)
- [Unsupervised and Distributional Detection of Machine-Generated Text](https://mlsanity.com/paper/2111.02878/) (2021)
- [Learning Invariant Representations of Social Media Users](https://mlsanity.com/paper/1910.04979/) (2019)
- [Low-Resource Authorship Style Transfer: Can Non-Famous Authors Be Imitated?](https://mlsanity.com/paper/2212.08986/) (2022)

## Built on by (most cited first, 30 of 40)

- [RAID: A Shared Benchmark for Robust Evaluation of Machine-Generated Text Detectors](https://mlsanity.com/paper/2405.07940/) (2024)
- [Detecting Multimedia Generated by Large AI Models: A Survey](https://mlsanity.com/paper/2402.00045/) (2024)
- [DeTeCtive: Detecting AI-generated Text via Multi-Level Contrastive Learning](https://mlsanity.com/paper/2410.20964/) (2024)
- [M4GT-Bench: Evaluation Benchmark for Black-Box Machine-Generated Text Detection](https://mlsanity.com/paper/2402.11175/) (2024)
- [Benchmarking Linguistic Diversity of Large Language Models](https://mlsanity.com/paper/2412.10271/) (2024)
- [Are We in the AI-Generated Text World Already? Quantifying and Monitoring AIGT on Social Media](https://mlsanity.com/paper/2412.18148/) (2024)
- [SemEval-2024 Task 8: Multidomain, Multimodel and Multilingual Machine-Generated Text Detection](https://mlsanity.com/paper/2404.14183/) (2024)
- [CollabStory: Multi-LLM Collaborative Story Generation and Authorship Analysis](https://mlsanity.com/paper/2406.12665/) (2024)
- [Detecting AI-Generated Text: Factors Influencing Detectability with Current Methods](https://mlsanity.com/paper/2406.15583/) (2024)
- [Are AI-Generated Text Detectors Robust to Adversarial Perturbations?](https://mlsanity.com/paper/2406.01179/) (2024)
- [DALD: Improving Logits-based Detector without Logits from Black-box LLMs](https://mlsanity.com/paper/2406.05232/) (2024)
- [DETree: DEtecting Human-AI Collaborative Texts via Tree-Structured Hierarchical Representation Learning](https://mlsanity.com/paper/2510.17489/) (2025)
- [Characterizing LLM-driven Social Network: The Chirper.ai Case](https://mlsanity.com/paper/2504.10286/) (2025)
- [Leveraging Multilingual Training for Authorship Representation: Enhancing Generalization across Languages and Domains](https://mlsanity.com/paper/2509.16531/) (2025)
- [Mitigating Paraphrase Attacks on Machine-Text Detectors via Paraphrase Inversion](https://mlsanity.com/paper/2410.21637/) (2024)
- [Synthetic Lyrics Detection Across Languages and Genres](https://mlsanity.com/paper/2406.15231/) (2024)
- [Who Writes What: Unveiling the Impact of Author Roles on AI-generated Text Detection](https://mlsanity.com/paper/2502.12611/) (2025)
- [Beyond the Final Actor: Modeling the Dual Roles of Creator and Editor for Fine-Grained LLM-Generated Text Detection](https://mlsanity.com/paper/2604.04932/) (2026)
- [Kill two birds with one stone: generalized and robust AI-generated text detection via dynamic perturbations](https://mlsanity.com/paper/2504.21019/) (2025)
- [AI-Generated Song Detection via Lyrics Transcripts](https://mlsanity.com/paper/2506.18488/) (2025)
- [Affect, Body, Cognition, Demographics, and Emotion: The ABCDE of Text Features for Computational Affective Science](https://mlsanity.com/paper/2512.17752/) (2025)
- [AICD Bench: A Challenging Benchmark for AI-Generated Code Detection](https://mlsanity.com/paper/2602.02079/) (2026)
- [Fine-Grained Bias Detection in LLM: Enhancing detection mechanisms for nuanced biases](https://mlsanity.com/paper/2503.06054/) (2025)
- [Towards Structurally Explainable Machine-Generated Text Detection: A Graph-Perspective Framework](https://mlsanity.com/paper/2505.12507/) (2025)
- [Attacks on Machine-Text Detectors Retain Stylistic Fingerprints](https://mlsanity.com/paper/2505.14608/) (2025)
- [Traceable TTS: Toward Watermark-Free TTS with Strong Traceability](https://mlsanity.com/paper/2507.03887/) (2025)
- [Two Birds with One Stone: Multi-Task Detection and Attribution of LLM-Generated Text](https://mlsanity.com/paper/2508.14190/) (2025)
- [Explainable Disentangled Representation Learning for Generalizable Authorship Attribution in the Era of Generative AI](https://mlsanity.com/paper/2604.21300/) (2026)
- [DSIPA: Detecting LLM-Generated Texts via Sentiment-Invariant Patterns Divergence Analysis](https://mlsanity.com/paper/2604.26328/) (2026)
- [MELD: Multi-Task Equilibrated Learning Detector for AI-Generated Text](https://mlsanity.com/paper/2605.06903/) (2026)

## Abstract (arXiv)

The advent of instruction-tuned language models that convincingly mimic human writing poses a significant risk of abuse. However, such abuse may be counteracted with the ability to detect whether a piece of text was composed by a language model rather than a human author. Some previous approaches to this problem have relied on supervised methods by training on corpora of confirmed human- and machine- written documents. Unfortunately, model under-specification poses an unavoidable challenge for neural network-based detectors, making them brittle in the face of data shifts, such as the release of newer language models producing still more fluent text than the models used to train the detectors. Other approaches require access to the models that may have generated a document in question, which is often impractical. In light of these challenges, we pursue a fundamentally different approach not relying on samples from language models of concern at training time. Instead, we propose to leverage representations of writing style estimated from human-authored text. Indeed, we find that features effective at distinguishing among human authors are also effective at distinguishing human from machine authors, including state-of-the-art large language models like Llama-2, ChatGPT, and GPT-4. Furthermore, given a handful of examples composed by each of several specific language models of interest, our approach affords the ability to predict which model generated a given document. The code and data to reproduce our experiments are available at https://github.com/LLNL/LUAR/tree/main/fewshot_iclr2024.