Showing cs.CRShow all
2 papers · 1 filter
cs.CR2026
CORE-BREW: LLR-Based Soft Decoding for Robust Multi-Bit LLM Watermarking
Joeun Kim, HoEun Kim, Young-Sik Kim
Reliable provenance for LLM outputs requires multi-bit watermarks that remain robust under editing while maintaining low false-positive rates. Existing ECC-based LLM watermarks rel…
cs.CR2026
Block-wise Codeword Embedding for Reliable Multi-bit Text Watermarking
Joeun Kim, HoEun Kim, Dongsup Jin +1
Recent multi-bit watermarking methods for large language models (LLMs) prioritize capacity over reliability, often conflating decoding with detection. Our analysis reveals that exi…