A lightweight entropy classifier plus an adaptive threshold method can watermark and detect low-entropy LLM code outputs without querying the original model, matching much larger detectors at 99% fewer detection-phase parameters.
Title resolution pending
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
-
Invisible Entropy: Towards Safe and Efficient Low-Entropy LLM Watermarking
A lightweight entropy classifier plus an adaptive threshold method can watermark and detect low-entropy LLM code outputs without querying the original model, matching much larger detectors at 99% fewer detection-phase parameters.