DERMARK dynamically segments LLM output by entropy to embed multi-bit watermarks more efficiently, though its theoretical derivation is unsound.
Can llm-generated misinformation be detected? In The Twelfth International Conference on Learning Representations, ICLR 2024, Vienna, Austria, May 7-11, 2024
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2025 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
DERMARK: A Dynamic, Efficient and Robust Multi-bit Watermark for Large Language Models
DERMARK dynamically segments LLM output by entropy to embed multi-bit watermarks more efficiently, though its theoretical derivation is unsound.