CoPA fools eight AI-text detectors by using a language model to paraphrase text while subtracting machine-like word probabilities during decoding, achieving high fooling rates without any training.
May contain 1-2 subtle non-native phrasings, but remains highly readable
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Your Language Model Can Secretly Write Like Humans: Contrastive Paraphrase Attacks on LLM-Generated Text Detectors
CoPA fools eight AI-text detectors by using a language model to paraphrase text while subtracting machine-like word probabilities during decoding, achieving high fooling rates without any training.