Dates: 2024
Principal Investigator: Dr. Aiman El-Maleh
Description: A complete approach to building and implementing a scalable solution for detecting model-generated text while maintaining information integrity in a variety of settings. All existing options will be investigated, and an efficient scalable solution will be proposed. The minimum number of tokens needed to recognize model-generated text will be determined. In addition, the suggested system will be able to recognize model-generated parts inside a text. The performance of the suggested approach in detecting model-generated text will be assessed using an examination of false positives created by the system.