GPT detectors are biased against non-native English writers
Explore this paper's citation graph
Summary
GPT detectors frequently misclassify non-native English writing as AI generated, raising concerns about fairness and robustness, and addressing the biases in these detectors is crucial to prevent the marginalization of non- natives in evaluative and educational settings.
- Type
- preprint
- Published
- 2023-04-06
- Cited by
- 650
- References
- 46
- Access
- Open access
- OpenAlex
- https://openalex.org/W4362706570
- Semantic Scholar
- https://api.semanticscholar.org/CorpusID:257985499
Keywords
Conversation, Detector, Robustness (evolution), Computer science, English language
References
- Automated Essay Scoring.
- Does writing development equal writing quality? A computational investigation of syntactic complexity in L2 learners
- Vocabulary size and use: Lexical richness in L2 written production
- Should we use characteristics of conversation to measure grammatical complexity in L2 writing development
- A Corpus-Based Evaluation of Syntactic Complexity Measures as Indices of College-Level ESL Writers' Language Development
- Short texts, best-fitting curves and new measures of lexical diversity
- Syntactic Complexity Measures and Their Relationship to L2 Proficiency: A Research Synthesis of College-Level L2 Writing.
- Lexical richness in the spontaneous speech of bilinguals
- GLTR: Statistical Detection and Visualization of Generated Text
- Release Strategies and the Social Impacts of Language Models
- Automatic Detection of Generated Text is Easiest when Humans are Fooled
- TweepFake: About detecting deepfake tweets
- Automatic Detection of Machine Generated Text: A Critical Survey
- All the News That’s Fit to Fabricate: AI-Generated Text as a Tool of Media Misinformation
- All That’s ‘Human’ Is Not Gold: Evaluating Human Evaluation of Generated Text
- Critical survey
- Evaluating Human-Language Model Interaction
- Comparing scientific abstracts generated by ChatGPT to original abstracts using an artificial intelligence output detector, plagiarism detector, and blinded human reviewers
- Abstracts written by ChatGPT fool scientists
- Tools such as ChatGPT threaten transparent science; here are our ground rules for their use
Cited by
- A real-world test of artificial intelligence infiltration of a university examinations system: A “Turing Test” case study
- A sustainable future
- DetectGPT: Zero-Shot Machine-Generated Text Detection using Probability Curvature
- Can AI-Generated Text be Reliably Detected?
- On the Possibilities of AI-Generated Text Detection
- Accuracy on the Curve: On the Nonlinear Correlation of ML Performance Between Data Subpopulations
- Watermarking Text Generated by Black-Box Language Models
- M4: Multi-generator, Multi-domain, and Multi-lingual Black-Box Machine-Generated Text Detection
- Ghostbuster: Detecting Text Ghostwritten by Large Language Models
- DNA-GPT: Divergent N-Gram Analysis for Training-Free Detection of GPT-Generated Text
- Challenges and Remedies to Privacy and Security in AIGC: Exploring the Potential of Privacy Computing, Blockchain, and Beyond
- On the Reliability of Watermarks for Large Language Models
- Intrinsic Dimension Estimation for Robust Detection of AI-Generated Texts
- Undetectable Watermarks for Language Models
- Provable Robust Watermarking for AI-Generated Text
- What Should Data Science Education Do with Large Language Models?
- Machine-Generated Text: A Comprehensive Survey of Threat Models and Detection Methods
- RADAR: Robust AI-Text Detection via Adversarial Learning
- Why technical solutions for detecting AI-generated content in research and education are insufficient
- Decoding Academic Integrity Policies: A Corpus Linguistics Investigation of AI and Other Technological Threats
Related papers
- The effect of familiarity of conversation partners on conversation turns contributed by augmented and typical speakers.
- Conversation and the Relevance and Turn Transition in Conversation
- 深まりのある意見交換は接觸場面において どのように達成されるか - 對話／共話との關連から -
- Study investigating the ease of talking via a robot tele-operated from same or different rooms
- Structure of Karonese Conversation in the Funeral: Case Study in Indonesia
- A Preliminary Study on Conversation Repair
- 深まりのある意見交換は接触場面において どのように達成されるか － 対話／共話との関連から －
- Teaching Speaking by Using Short Conversation
- Effects of conversation on intracranial pressure in comatose patients.