mlsanityData to 7 Oct 2026

Technical Report on the Pangram AI-Generated Text Classifier

arXiv 2402.14873 ↗First posted 21 Feb 2024cs.CL cs.AIcited by 32

Not on the map. Why?

Abstract (arXiv)

We present Pangram Text, a transformer-based neural network trained to distinguish text written by large language models from text written by humans. Pangram Text outperforms zero-shot methods such as DetectGPT as well as leading commercial AI detection tools with over 38 times lower error rates on a comprehensive benchmark comprised of 10 text domains (student writing, creative writing, scientific writing, books, encyclopedias, news, email, scientific papers, short-form Q&A) and 8 open- and closed-source large language models. We propose a training algorithm, hard negative mining with synthetic mirrors, that enables our classifier to achieve orders of magnitude lower false positive rates on high-data domains such as reviews. Finally, we show that Pangram Text is not biased against nonnative English speakers and generalizes to domains and models unseen during training.

Key facts. Cited by 32 papers in the map (data through 7 Oct 2026). Builds on 16 of its 34 references that are in the map. First posted on arXiv 21 Feb 2024.

Builds on

16 of 34 references.

References · 16

  1. 2017Attention Is All You Needcited by 59,932
  2. 2018Hierarchical Neural Story Generationcited by 1,182
  3. 2023DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvaturecited by 705
  4. 2016Training Region-based Object Detectors with Online Hard Example Miningcited by 577
  5. 2020A Survey of Deep Active Learningcited by 547
  6. 2019GLTR: Statistical Detection and Visualization of Generated Textcited by 490
  7. 2023Paraphrasing evades detectors of AI-generated text, but retrieval is an effective defensecited by 401
  8. 2021XL-Sum: Large-Scale Multilingual Abstractive Summarization for 44 Languagescited by 346
  9. 2023DetectLLM: Leveraging Log Rank Information for Zero-Shot Detection of Machine-Generated Textcited by 207
  10. 2020The Multilingual Amazon Reviews Corpuscited by 142
  11. 2019A Question-Entailment Approach to Question Answeringcited by 133
  12. 2023Testing of Detection Tools for AI-Generated Textcited by 86
4 more
  1. 2019Active Learning for Deep Detection Neural Networkscited by 62
  2. 2023Smaller Language Models are Better Black-box Machine-Generated Text Detectorscited by 35
  3. 2023Counter Turing Test CT^2: AI-Generated Text Detection is Not as Easy as You May Think -- Introducing AI Detectability Indexcited by 14
  4. 2023On the Zero-Shot Generalization of Machine-Generated Text Detectorscited by 11

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 2; 2025: 10; 2026: 20.

32 papers.

Sort:

By area:

  1. 2025Flipping the Dialogue: Training and Evaluating User Language Modelscited by 50
  2. 2025People who frequently use ChatGPT for writing tasks are accurate and robust detectors of AI-generated textcited by 39
  3. 2025EditLens: Quantifying the Extent of AI Editing in Textcited by 26
  4. 2025AI-Slop to AI-Polish? Aligning Language Models through Edit-Based Writing Rewards and Test-time Computationcited by 24
  5. 2025DAMAGE: Detecting Adversarially Modified AI Generated Textcited by 20
  6. 2026GPTZero: Robust Detection of LLM-Generated Textscited by 10
  7. 2025AI use in American newspapers is widespread, uneven, and rarely disclosedcited by 9
  8. 2026StoryScope: Investigating idiosyncrasies in AI fictioncited by 9
  9. 2025Frankentext: Stitching random text fragments into long-form narrativescited by 8
  10. 2024ESPERANTO: Evaluating Synthesized Phrases to Enhance Robustness in AI Detection for Text Originationcited by 6
  11. 2025Death of the Novel(ty): Beyond n-Gram Novelty as a Metric for Textual Creativitycited by 4
  12. 2026LLM-Generated or Human-Written? Comparing Review and Non-Review Papers on ArXivcited by 3
20 more
  1. 2026'Your AI Text is not Mine': Redefining and Evaluating AI-generated Text Detection under Realistic Assumptionscited by 3
  2. 2026Pangram 4 Technical Reportcited by 3
  3. 2026Base Models Look Human To AI Detectorscited by 2
  4. 2026Gaming AI-Assisted Peer Reviews Poses New Risks to the Scientific Communitycited by 2
  5. 2026Black-Box Forensics for Conversational LLM Agentscited by 2
  6. 2025ExaGPT: Example-Based Machine-Generated Text Detection for Human Interpretabilitycited by 1
  7. 2026DoPE: Decoy Oriented Perturbation Encapsulation Human-Readable, AI-Hostile Documents for Academic Integritycited by 1
  8. 2026Can You Make It Sound Like You? Post-Editing LLM-Generated Text for Personal Stylecited by 1
  9. 2026Hitting a Moving Target: Test-Time Adaptation for AI Text Detection under Continual Distribution Shiftcited by 1
  10. 2024Gram2Vec: An Interpretable Document Vectorizercited by 0
  11. 2025Domain Gating Ensemble Networks for AI-Generated Text Detectioncited by 0
  12. 2026Integrity Shield A System for Ethical AI Use & Authorship Transparency in Assessmentscited by 0
  13. 2026Show, Don't TELL: Explainable AI-Generated Text Detectioncited by 0
  14. 2026Measuring, Localizing, and Ablating Alignment Signatures in LLMscited by 0
  15. 2026Non-Parametric Machine Text Detection via Multi-View Gaussian Processescited by 0
  16. 2026Team DACTYL at PAN 2026: Bayesian Data Mixing and Empirical X-risk Minimization for AI-text Detectioncited by 0
  17. 2026Stability-Aware Feature Design for Robust Watermark Detection in Machine-Generated Textcited by 0
  18. 2026Pretrained Persona Mixture Models and Tandem Models for Human Simulationcited by 0
  19. 2026Science or Slop?: Benchmarking and Mitigating Scientific Slop in AI-Generated Paperscited by 0
  20. 2026Large Language Model Turnover Undermines Screening for Artificial Intelligence-Assisted Scientific Writingcited by 0