Technical Report on the Pangram AI-Generated Text Classifier

• URL: http://arxiv.org/abs/2402.14873v3
• Date: Mon, 29 Jul 2024 08:27:34 GMT
• Title: Technical Report on the Pangram AI-Generated Text Classifier
• Authors: Bradley Emi, Max Spero,
• Abstract summary: We present Pangram Text, a transformer-based neural network trained to distinguish text written by large language models from text written by humans. We show that Pangram Text is not biased against nonnative English speakers and generalizes to domains and models unseen during training.
• Score: 0.14732811715354457
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: 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.

Related papers

• Deep Learning Detection Method for Large Language Models-Generated Scientific Content [0.0]
  Large Language Models generate scientific content that is indistinguishable from that written by humans. This research paper presents a novel ChatGPT-generated scientific text detection method, AI-Catcher. On average, AI-Catcher improved accuracy by 37.4%.
  arXiv  Detail & Related papers  (2024-02-27T19:16:39Z)
• Offline Detection of Misspelled Handwritten Words by Convolving Recognition Model Features with Text Labels [0.0]
  We introduce the task of comparing a handwriting image to text. Our model's classification head is trained entirely on synthetic data created using a state-of-the-art generative adversarial network. Such massive performance gains can lead to significant productivity increases in applications utilizing human-in-the-loop automation.
  arXiv  Detail & Related papers  (2023-09-18T21:13:42Z)
• Distinguishing Human Generated Text From ChatGPT Generated Text Using Machine Learning [0.251657752676152]
  This paper presents a machine learning-based solution that can identify the ChatGPT delivered text from the human written text. We have tested the proposed model on a Kaggle dataset consisting of 10,000 texts out of which 5,204 texts were written by humans and collected from news and social media. On the corpus generated by GPT-3.5, the proposed algorithm presents an accuracy of 77%.
  arXiv  Detail & Related papers  (2023-05-26T09:27:43Z)
• Smaller Language Models are Better Black-box Machine-Generated Text Detectors [56.36291277897995]
  Small and partially-trained models are better universal text detectors. We find that whether the detector and generator were trained on the same data is not critically important to the detection success. For instance, the OPT-125M model has an AUC of 0.81 in detecting ChatGPT generations, whereas a larger model from the GPT family, GPTJ-6B, has AUC of 0.45.
  arXiv  Detail & Related papers  (2023-05-17T00:09:08Z)
• Paraphrasing evades detectors of AI-generated text, but retrieval is an effective defense [56.077252790310176]
  We present a paraphrase generation model (DIPPER) that can paraphrase paragraphs, condition on surrounding context, and control lexical diversity and content reordering. Using DIPPER to paraphrase text generated by three large language models (including GPT3.5-davinci-003) successfully evades several detectors, including watermarking. We introduce a simple defense that relies on retrieving semantically-similar generations and must be maintained by a language model API provider.
  arXiv  Detail & Related papers  (2023-03-23T16:29:27Z)
• Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding [53.170767750244366]
  Imagen is a text-to-image diffusion model with an unprecedented degree of photorealism and a deep level of language understanding. To assess text-to-image models in greater depth, we introduce DrawBench, a comprehensive and challenging benchmark for text-to-image models.
  arXiv  Detail & Related papers  (2022-05-23T17:42:53Z)
• Open Vocabulary Electroencephalography-To-Text Decoding and Zero-shot Sentiment Classification [78.120927891455]
  State-of-the-art brain-to-text systems have achieved great success in decoding language directly from brain signals using neural networks. In this paper, we extend the problem to open vocabulary Electroencephalography(EEG)-To-Text Sequence-To-Sequence decoding and zero-shot sentence sentiment classification on natural reading tasks. Our model achieves a 40.1% BLEU-1 score on EEG-To-Text decoding and a 55.6% F1 score on zero-shot EEG-based ternary sentiment classification, which significantly outperforms supervised baselines.
  arXiv  Detail & Related papers  (2021-12-05T21:57:22Z)
• How much do language models copy from their training data? Evaluating linguistic novelty in text generation using RAVEN [63.79300884115027]
  Current language models can generate high-quality text. Are they simply copying text they have seen before, or have they learned generalizable linguistic abstractions? We introduce RAVEN, a suite of analyses for assessing the novelty of generated text.
  arXiv  Detail & Related papers  (2021-11-18T04:07:09Z)
• Lattice-BERT: Leveraging Multi-Granularity Representations in Chinese Pre-trained Language Models [62.41139712595334]
  We propose a novel pre-training paradigm for Chinese -- Lattice-BERT. We construct a lattice graph from the characters and words in a sentence and feed all these text units into transformers. We show that our model can bring an average increase of 1.5% under the 12-layer setting.
  arXiv  Detail & Related papers  (2021-04-15T02:36:49Z)
• Offline Handwritten Chinese Text Recognition with Convolutional Neural Networks [5.984124397831814]
  In this paper, we build the models using only the convolutional neural networks and use CTC as the loss function. We achieve 6.81% character error rate (CER) on the ICDAR 2013 competition set, which is the best published result without language model correction.
  arXiv  Detail & Related papers  (2020-06-28T14:34:38Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.