Unverified50% confidenceCautionExact time
Simple editing operations such as replacing synonyms, adjusting word order, and inserting or deleting words can also weaken watermarks, and modifying a sufficient proportion of tokens can bring the green-list statistics back to a random level.
1
Sources
50%
Confidence
Long-term
Relevance
9/11/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedInvisible watermarking technology embeds statistical signals during the generation phase of large language models by fine-tuning the token sampling strategy (dividing the vocabulary into green and red tokens and biasing toward green tokens).70% similarUnverifiedFixed-character chunking strategies truncate sentences and disrupt semantic integrity, and should be replaced with intelligent semantic chunking or sliding window strategies.68% similarUnverifiedThe green token watermarking method divides the vocabulary into green and red groups based on the hash value of the previous token before sampling, applies probability weighting to green tokens, and causes the frequency of green tokens in watermarked text to be significantly higher than the natural language baseline68% similarUnverifiedIntroducing randomized paraphrasing or standardizing outputs can disrupt hidden encodings and strip out smuggled information.67% similarUnverifiedPrefix caching improves inference efficiency by caching common prompt prefixes to reduce redundant computation.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/898148API
curl https://kongchang.com/api/v1/knowledge/claims/898148MCP
get_claim(id=898148)