# RepreGuard: Detecting LLM-Generated Text by Revealing Hidden Representation Patterns

arXiv: https://arxiv.org/abs/2508.13152 · HTML: https://mlsanity.com/paper/2508.13152/

## Key facts

- Cited by 24 papers in the map (data through 14 Oct 2026).
- Builds on 40 of its 55 references that are in the map.
- First posted on arXiv 18 Aug 2025.

## Builds on (most cited first, 30 of 40)

- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [RoBERTa: A Robustly Optimized BERT Pretraining Approach](https://mlsanity.com/paper/1907.11692/) (2019)
- [The Curious Case of Neural Text Degeneration](https://mlsanity.com/paper/1904.09751/) (2019)
- [Character-level Convolutional Networks for Text Classification](https://mlsanity.com/paper/1509.01626/) (2015)
- [Extracting Training Data from Large Language Models](https://mlsanity.com/paper/2012.07805/) (2020)
- [Survey of Hallucination in Natural Language Generation](https://mlsanity.com/paper/2202.03629/) (2022)
- [Representation Engineering: A Top-Down Approach to AI Transparency](https://mlsanity.com/paper/2310.01405/) (2023)
- [Don't Give Me the Details, Just the Summary! Topic-Aware Convolutional Neural Networks for Extreme Summarization](https://mlsanity.com/paper/1808.08745/) (2018)
- [Hierarchical Neural Story Generation](https://mlsanity.com/paper/1805.04833/) (2018)
- [WizardLM: Empowering large pre-trained language models to follow complex instructions](https://mlsanity.com/paper/2304.12244/) (2023)
- [DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvature](https://mlsanity.com/paper/2301.11305/) (2023)
- [The Internal State of an LLM Knows When It's Lying](https://mlsanity.com/paper/2304.13734/) (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)
- [How Close is ChatGPT to Human Experts? Comparison Corpus, Evaluation, and Detection](https://mlsanity.com/paper/2301.07597/) (2023)
- [Paraphrasing evades detectors of AI-generated text, but retrieval is an effective defense](https://mlsanity.com/paper/2303.13408/) (2023)
- [TextBugger: Generating Adversarial Text Against Real-world Applications](https://mlsanity.com/paper/1812.05271/) (2018)
- [Orca: Progressive Learning from Complex Explanation Traces of GPT-4](https://mlsanity.com/paper/2306.02707/) (2023)
- [Fast-DetectGPT: Efficient Zero-Shot Detection of Machine-Generated Text via Conditional Probability Curvature](https://mlsanity.com/paper/2310.05130/) (2023)
- [Spotting LLMs With Binoculars: Zero-Shot Detection of Machine-Generated Text](https://mlsanity.com/paper/2401.12070/) (2024)
- [DetectLLM: Leveraging Log Rank Information for Zero-Shot Detection of Machine-Generated Text](https://mlsanity.com/paper/2306.05540/) (2023)
- [Do LLMs exhibit human-like response biases? A case study in survey design](https://mlsanity.com/paper/2311.04076/) (2023)
- [RAID: A Shared Benchmark for Robust Evaluation of Machine-Generated Text Detectors](https://mlsanity.com/paper/2405.07940/) (2024)
- [DNA-GPT: Divergent N-Gram Analysis for Training-Free Detection of GPT-Generated Text](https://mlsanity.com/paper/2305.17359/) (2023)
- [Language-Specific Neurons: The Key to Multilingual Capabilities in Large Language Models](https://mlsanity.com/paper/2402.16438/) (2024)
- [TweepFake: about Detecting Deepfake Tweets](https://mlsanity.com/paper/2008.00036/) (2020)
- [A Survey on LLM-Generated Text Detection: Necessity, Methods, and Future Directions](https://mlsanity.com/paper/2310.14724/) (2023)
- [Analyzing Individual Neurons in Pre-trained Language Models](https://mlsanity.com/paper/2010.02695/) (2020)
- [Ghostbuster: Detecting Text Ghostwritten by Large Language Models](https://mlsanity.com/paper/2305.15047/) (2023)
- [Neurons in Large Language Models: Dead, N-gram, Positional](https://mlsanity.com/paper/2309.04827/) (2023)

## Built on by (most cited first, 24 of 24)

- [Benchmarking the Detection of LLMs-Generated Modern Chinese Poetry](https://mlsanity.com/paper/2509.01620/) (2025)
- [Learn-to-Distance: Distance Learning for Detecting LLM-Generated Text](https://mlsanity.com/paper/2601.21895/) (2026)
- [Controlling Repetition in Protein Language Models](https://mlsanity.com/paper/2602.00782/) (2026)
- [DetectRL-X: Towards Reliable Multilingual and Real-World LLM-Generated Text Detection](https://mlsanity.com/paper/2605.15518/) (2026)
- [Beyond Raw Detection Scores: Markov-Informed Calibration for Boosting Machine-Generated Text Detection](https://mlsanity.com/paper/2602.08031/) (2026)
- [Word Recovery in Large Language Models Enables Character-Level Tokenization Robustness](https://mlsanity.com/paper/2603.10771/) (2026)
- [WaveDetect: Robust Framework for Machine-Generated Text Detection via Wavelet Transform](https://mlsanity.com/paper/2606.23336/) (2026)
- [Faithful-Patchscopes: Understanding and Mitigating Model Bias in Hidden Representations Explanation of Large Language Models](https://mlsanity.com/paper/2602.00300/) (2026)
- [MELD: Multi-Task Equilibrated Learning Detector for AI-Generated Text](https://mlsanity.com/paper/2605.06903/) (2026)
- [Seeing the Poem: Image-Semantic Detection of AI-Generated Modern Chinese Poetry with MLLMs](https://mlsanity.com/paper/2605.22654/) (2026)
- [EVIL-Detect for NLPCC 2026 Shared Task 6: LLM-Generated Text Detection](https://mlsanity.com/paper/2608.10698/) (2026)
- [Exons-Detect: Identifying and Amplifying Exonic Tokens via Hidden-State Discrepancy for Robust AI-Generated Text Detection](https://mlsanity.com/paper/2603.24981/) (2026)
- [Log-Likelihood, Simpson's Paradox, and the Detection of Machine-Generated Text](https://mlsanity.com/paper/2605.06294/) (2026)
- [Steer-to-Detect: Probing Hidden Representations for Detection of LLM-Generated Texts](https://mlsanity.com/paper/2605.12890/) (2026)
- [Multi-Level Contextual Token Relation Modeling for Machine-Generated Text Detection](https://mlsanity.com/paper/2605.16107/) (2026)
- [Hidden Human-Like Nature of Machine-Generated Texts: Theory and Detection Enhancement](https://mlsanity.com/paper/2605.23190/) (2026)
- [LLM Self-Recognition: Steering and Retrieving Activation Signatures](https://mlsanity.com/paper/2606.06315/) (2026)
- [SV-Detect: AI-generated Text Detection with Steering Vectors](https://mlsanity.com/paper/2606.07313/) (2026)
- [RepBench: Compiling Benchmarks into Capability Representations for Large Language Models](https://mlsanity.com/paper/2607.28008/) (2026)
- [Build it, Break it, Repeat: Benchmarking and improving LLM-manipulated disinformation detection in social media posts](https://mlsanity.com/paper/2608.09510/) (2026)
- [Linear Probing Provides Robust and Efficient Detection of Machine-Generated Text](https://mlsanity.com/paper/2608.24780/) (2026)
- [IndicDetect: Evaluating Cross-Lingual LLM-Generated Text Detection for Hindi, Telugu, and Tamil](https://mlsanity.com/paper/2608.29919/) (2026)
- [Neuron-Guided Fine-Tuning: Unlocking Efficient Alignment Mechanisms for Large Language Models](https://mlsanity.com/paper/2609.05913/) (2026)
- [DeBERTa-ConPara: Attack-Aware and Deployment-Realistic Detection of AI-Generated Text](https://mlsanity.com/paper/2610.00883/) (2026)

## Abstract (arXiv)

Detecting content generated by large language models (LLMs) is crucial for preventing misuse and building trustworthy AI systems. Although existing detection methods perform well, their robustness in out-of-distribution (OOD) scenarios is still lacking. In this paper, we hypothesize that, compared to features used by existing detection methods, the internal representations of LLMs contain more comprehensive and raw features that can more effectively capture and distinguish the statistical pattern differences between LLM-generated texts (LGT) and human-written texts (HWT). We validated this hypothesis across different LLMs and observed significant differences in neural activation patterns when processing these two types of texts. Based on this, we propose RepreGuard, an efficient statistics-based detection method. Specifically, we first employ a surrogate model to collect representation of LGT and HWT, and extract the distinct activation feature that can better identify LGT. We can classify the text by calculating the projection score of the text representations along this feature direction and comparing with a precomputed threshold. Experimental results show that RepreGuard outperforms all baselines with average 94.92% AUROC on both in-distribution (ID) and OOD scenarios, while also demonstrating robust resilience to various text sizes and mainstream attacks. Data and code are publicly available at: https://github.com/NLP2CT/RepreGuard