Skip to main navigation
Skip to search
Skip to main content
Hong Kong Baptist University Home
Help & FAQ
Home
Scholars
Departments / Units
Research Output
Projects / Grants
Prizes / Awards
Activities
Press/Media
Student theses
Datasets
Search by expertise, name or affiliation
RepreGuard: Detecting LLM-Generated Text by Revealing Hidden Representation Patterns
Xin Chen (Co-first author)
, Junchao Wu (Co-first author)
, Shu Yang
, Runzhe Zhan
, Zeyu Wu
, Ziyang Luo
, Di Wang
, Min Yang
, Lidia S. Chao
, Derek F. Wong
*
*
Corresponding author for this work
Department of Computer Science
Research output
:
Contribution to journal
›
Journal article
›
peer-review
Overview
Fingerprint
Fingerprint
Dive into the research topics of 'RepreGuard: Detecting LLM-Generated Text by Revealing Hidden Representation Patterns'. Together they form a unique fingerprint.
Sort by:
Weight
Alphabetically
Keyphrases
Large Language Models
100%
Hidden Representation
100%
Computer-generated Text
100%
Detection Method
60%
Written Text
40%
Statistical Methods
20%
Resilience
20%
Statistical Pattern
20%
Text Representation
20%
Text Types
20%
AI-based Methods
20%
Surrogate Model
20%
Text Size
20%
In-distribution
20%
Area under the Receiver Operating Characteristic (AUROC)
20%
Trustworthy AI
20%
Neural Activation Patterns
20%
Internal Representation
20%
Activation Feature
20%
Computer Science
Large Language Model
100%
Detection Method
42%
Experimental Result
14%
Text Representation
14%
Generated Content
14%
Activation Pattern
14%
Difference Pattern
14%
Trustworthy AI
14%
Internal Representation
14%