Deciphering Textual Authenticity: A Generalized Strategy through the Lens of Large Language Semantics for Detecting Human vs. Machine-Generated Text.
Mazal BethanyBrandon WherryEmet BethanyNishant VishwamitraAnthony RiosPeyman NajafiradPublished in: USENIX Security Symposium (2024)
Keyphrases
- human generated
- logical language
- specification language
- human language
- operational semantics
- natural language
- textual data
- language generation
- intended meaning
- logic programming language
- conceptual graphs
- formal language
- model theoretic
- semantic structure
- programming language
- highly expressive
- conditional logic
- formal semantics
- english text
- manually constructed
- keywords
- digital images
- text to speech synthesis
- proof theory
- semantic information
- action language
- dynamic logic
- definition language
- formal languages
- modal operators
- textual features
- fixed point semantics
- text content
- english language
- semantic representations
- free text
- word meanings
- native language
- speech acts
- language learning
- conceptual model
- logic programming
- relational structures
- text mining
- logic programs
- object oriented
- language acquisition
- textual case based reasoning
- plain text
- computational linguistics