Lune

CCS2024Top-tier venue

MGTBench: Benchmarking Machine-Generated Text Detection

Xinlei He, Xinyue Shen, Zeyuan Chen, Michael Backes, Yang Zhang

Abstract

Nowadays, powerful large language models (LLMs) such as ChatGPT have demonstrated revolutionary power in a variety of natural language processing (NLP) tasks such as text classification, sentiment analysis, language translation, and question-answering. Consequently, the detection of machine-generated texts (MGTs) is becoming increasingly crucial as LLMs become more advanced and prevalent. These models have the ability to generate human-like language, making it challenging to discern whether a text is authored by a human or a machine. This raises concerns regarding authenticity, accountability, and potential bias. However, existing methods for detecting MGTs are evaluated using different model architectures, datasets, and experimental settings, resulting in a lack of a comprehensive evaluation framework that encompasses various methodologies. Furthermore, it remains unclear how existing detection methods would perform against powerful LLMs. In this paper, we fill this gap by proposing the first benchmark framework for MGT detection against powerful LLMs, named MGTBench. Extensive evaluations on public datasets with curated texts generated by various powerful LLMs such as ChatGPT-turbo and Claude demonstrate the effectiveness of different detection methods. Our ablation study shows that a larger number of words in general leads to better performance and most detection methods can achieve similar performance with much fewer training samples. Additionally, our findings reveal that metric-based/model-based detection methods exhibit better transferability across different LLMs/datasets. Furthermore, we delve into a more challenging task: text attribution, where the goal is to identify the originating model of a given text, i.e., whether it is a specific LLM or authored by a human. Our findings indicate that the model-based detection methods still perform well in the text attribution task. To investigate the robustness of different detection methods, we consider three adversarial attacks, namely paraphrasing, random spacing, and adversarial perturbations. We discover that these attacks can significantly diminish detection effectiveness, underscoring the critical need for the development of more robust detection methods. We envision that MGTBench will serve as a benchmark tool to accelerate future investigations involving the evaluation of powerful MGT detection methods on their respective datasets and the development of more advanced MGT detection methods. 1

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

41 top-tier papers have cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • BiScope: AI-generated Text Detection by Checking Memorization of Preceding Tokens cites DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvature
  • BiScope: AI-generated Text Detection by Checking Memorization of Preceding Tokens cites Intrinsic Dimension Estimation for Robust Detection of AI-Generated Texts
  • Beyond Binary: Towards Fine-Grained LLM-Generated Text Detection via Role Recognition and Involvement Measurement cites DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvature
  • Beyond Binary: Towards Fine-Grained LLM-Generated Text Detection via Role Recognition and Involvement Measurement cites Authorship Attribution for Neural Text Generation
  • DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvature cites Language Models are Few-Shot Learners
  • DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvature cites Authorship Attribution for Neural Text Generation
  • TURINGBENCH: A Benchmark Environment for Turing Test in the Age of Neural Text Generation cites Language Models are Few-Shot Learners
  • TURINGBENCH: A Benchmark Environment for Turing Test in the Age of Neural Text Generation cites Authorship Attribution for Neural Text Generation
  • Intrinsic Dimension Estimation for Robust Detection of AI-Generated Texts cites DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvature
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers41

  1. Intrinsic Dimension Estimation for Robust Detection of AI-Generated TextsEduard Tulchinskii, Kristian Kuznetsov, Laida Kushnareva, Daniil Cherniavskii et al.NeurIPS 2023 · 163 citations
  2. "Do Anything Now": Characterizing and Evaluating In-The-Wild Jailbreak Prompts on Large Language ModelsXinyue Shen, Zeyuan Chen, Michael Backes, Yun Shen et al.CCS 2024 · 132 citations
  3. BiScope: AI-generated Text Detection by Checking Memorization of Preceding TokensHanxi Guo, Siyuan Cheng, Xiaolong Jin, Zhuo Zhang et al.NeurIPS 2024 · 52 citations
  4. Where Did I Come From? Origin Attribution of AI-Generated ImagesZhenting Wang, Chen Chen, Yi Zeng, Lingjuan Lyu et al.NeurIPS 2023 · 44 citations
  5. Beyond Binary: Towards Fine-Grained LLM-Generated Text Detection via Role Recognition and Involvement MeasurementZihao Cheng, Li Zhou, Feng Jiang, Benyou Wang et al.WWW 2025 · 20 citations

Builds on8

  1. Language Models are Few-Shot LearnersTom B. Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah et al.NeurIPS 2020 · 64,255 citations
  2. TruthfulQA: Measuring How Models Mimic Human FalsehoodsStephanie Lin, Jacob Hilton, Owain EvansACL 2022 · 3,228 citations
  3. DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability CurvatureEric Mitchell, Yoonho Lee, Alexander Khazatsky, Christopher D. Manning et al.ICML 2023 · 988 citations
  4. Authorship Attribution for Neural Text GenerationAdaku Uchendu, Thai Le, Kai Shu, Dongwon LeeEMNLP 2020 · 110 citations
  5. TURINGBENCH: A Benchmark Environment for Turing Test in the Age of Neural Text GenerationAdaku Uchendu, Zeyu Ma, Thai Le, Rui Zhang et al.EMNLP Findings 2021 · 82 citations

Related papers

  1. M4GT-Bench: Evaluation Benchmark for Black-Box Machine-Generated Text DetectionYuxia Wang, Jonibek Mansurov, Petar Ivanov, Jinyan Su et al.ACL 2024
  2. OpenTuringBench: An Open-Model-based Benchmark and Framework for Machine-Generated Text Detection and AttributionLucio La Cava, Andrea TagarelliEMNLP 2025
  3. Stress-testing Machine Generated Text Detection: Shifting Language Models Writing Style to Fool DetectorsAndrea Pedrotti, Michele Papucci, Cristiano Ciaccio, Alessio Miaschi et al.ACL Findings 2025 · 2 citations
  4. DEMASQ: Unmasking the ChatGPT WordsmithKavita Kumari, Alessandro Pegoraro, Hossein Fereidooni, Ahmad-Reza SadeghiNDSS 2024
  5. Stumbling Blocks: Stress Testing the Robustness of Machine-Generated Text Detectors Under AttacksYichen Wang, Shangbin Feng, Abe Bohan Hou, Xiao Pu et al.ACL 2024
Dusk over the sea between two cliffs drawn in fine vertical lines