ChatGPT or Human? Detect and Explain. Explaining Decisions of Machine
Learning Model for Detecting Short ChatGPT-generated Text
- URL: http://arxiv.org/abs/2301.13852v1
- Date: Mon, 30 Jan 2023 08:06:08 GMT
- Title: ChatGPT or Human? Detect and Explain. Explaining Decisions of Machine
Learning Model for Detecting Short ChatGPT-generated Text
- Authors: Sandra Mitrovi\'c, Davide Andreoletti, Omran Ayoub
- Abstract summary: We study whether a machine learning model can be effectively trained to accurately distinguish between original human and seemingly human (that is, ChatGPT-generated) text.
We employ an explainable artificial intelligence framework to gain insight into the reasoning behind the model trained to differentiate between ChatGPT-generated and human-generated text.
Our study focuses on short online reviews, conducting two experiments comparing human-generated and ChatGPT-generated text.
- Score: 2.0378492681344493
- License: http://creativecommons.org/publicdomain/zero/1.0/
- Abstract: ChatGPT has the ability to generate grammatically flawless and
seemingly-human replies to different types of questions from various domains.
The number of its users and of its applications is growing at an unprecedented
rate. Unfortunately, use and abuse come hand in hand. In this paper, we study
whether a machine learning model can be effectively trained to accurately
distinguish between original human and seemingly human (that is,
ChatGPT-generated) text, especially when this text is short. Furthermore, we
employ an explainable artificial intelligence framework to gain insight into
the reasoning behind the model trained to differentiate between
ChatGPT-generated and human-generated text. The goal is to analyze model's
decisions and determine if any specific patterns or characteristics can be
identified. Our study focuses on short online reviews, conducting two
experiments comparing human-generated and ChatGPT-generated text. The first
experiment involves ChatGPT text generated from custom queries, while the
second experiment involves text generated by rephrasing original
human-generated reviews. We fine-tune a Transformer-based model and use it to
make predictions, which are then explained using SHAP. We compare our model
with a perplexity score-based approach and find that disambiguation between
human and ChatGPT-generated reviews is more challenging for the ML model when
using rephrased text. However, our proposed approach still achieves an accuracy
of 79%. Using explainability, we observe that ChatGPT's writing is polite,
without specific details, using fancy and atypical vocabulary, impersonal, and
typically it does not express feelings.
Related papers
- DEMASQ: Unmasking the ChatGPT Wordsmith [63.8746084667206]
We propose an effective ChatGPT detector named DEMASQ, which accurately identifies ChatGPT-generated content.
Our method addresses two critical factors: (i) the distinct biases in text composition observed in human- and machine-generated content and (ii) the alterations made by humans to evade previous detection methods.
arXiv Detail & Related papers (2023-11-08T21:13:05Z) - Primacy Effect of ChatGPT [69.49920102917598]
We study the primacy effect of ChatGPT: the tendency of selecting the labels at earlier positions as the answer.
We hope that our experiments and analyses provide additional insights into building more reliable ChatGPT-based solutions.
arXiv Detail & Related papers (2023-10-20T00:37:28Z) - Detecting ChatGPT: A Survey of the State of Detecting ChatGPT-Generated
Text [1.9643748953805937]
generative language models can potentially deceive by generating artificial text that appears to be human-generated.
This survey provides an overview of the current approaches employed to differentiate between texts generated by humans and ChatGPT.
arXiv Detail & Related papers (2023-09-14T13:05:20Z) - Is ChatGPT Involved in Texts? Measure the Polish Ratio to Detect
ChatGPT-Generated Text [48.36706154871577]
We introduce a novel dataset termed HPPT (ChatGPT-polished academic abstracts)
It diverges from extant corpora by comprising pairs of human-written and ChatGPT-polished abstracts instead of purely ChatGPT-generated texts.
We also propose the "Polish Ratio" method, an innovative measure of the degree of modification made by ChatGPT compared to the original human-written text.
arXiv Detail & Related papers (2023-07-21T06:38:37Z) - ChatGPT vs Human-authored Text: Insights into Controllable Text
Summarization and Sentence Style Transfer [8.64514166615844]
We conduct a systematic inspection of ChatGPT's performance in two controllable generation tasks.
We evaluate the faithfulness of the generated text, and compare the model's performance with human-authored texts.
We observe that ChatGPT sometimes incorporates factual errors or hallucinations when adapting the text to suit a specific style.
arXiv Detail & Related papers (2023-06-13T14:21:35Z) - 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) - GPT-Sentinel: Distinguishing Human and ChatGPT Generated Content [27.901155229342375]
We present a novel approach for detecting ChatGPT-generated vs. human-written text using language models.
Our models achieved remarkable results, with an accuracy of over 97% on the test dataset, as evaluated through various metrics.
arXiv Detail & Related papers (2023-05-13T17:12:11Z) - AI, write an essay for me: A large-scale comparison of human-written
versus ChatGPT-generated essays [66.36541161082856]
ChatGPT and similar generative AI models have attracted hundreds of millions of users.
This study compares human-written versus ChatGPT-generated argumentative student essays.
arXiv Detail & Related papers (2023-04-24T12:58:28Z) - To ChatGPT, or not to ChatGPT: That is the question! [78.407861566006]
This study provides a comprehensive and contemporary assessment of the most recent techniques in ChatGPT detection.
We have curated a benchmark dataset consisting of prompts from ChatGPT and humans, including diverse questions from medical, open Q&A, and finance domains.
Our evaluation results demonstrate that none of the existing methods can effectively detect ChatGPT-generated content.
arXiv Detail & Related papers (2023-04-04T03:04:28Z) - Real or Fake Text?: Investigating Human Ability to Detect Boundaries
Between Human-Written and Machine-Generated Text [23.622347443796183]
We study a more realistic setting where text begins as human-written and transitions to being generated by state-of-the-art neural language models.
We show that, while annotators often struggle at this task, there is substantial variance in annotator skill and that given proper incentives, annotators can improve at this task over time.
arXiv Detail & Related papers (2022-12-24T06:40:25Z)
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.