Anthropic 解释 Claude 的不可见文本水印将如何运作

以这句话为例：“今天天气很冷，而且……”。下一个词极不可能是“含糖的”。但“阴天”或“灰暗”则很有可能。在大多数情况下，模型最终选择这两个词中的哪一个，对读者而言并没有太大影响——无论哪种选择，句子的含义基本相同。在这类情况下，具体选择通常由一个随机数决定。

水印技术会利用这类低风险选择——它们会在一段生成文本中多次出现——在 Claude 的回复中留下某种模式。这种模式对读者不可见，但任何拥有能够编码该模式的密钥的人都可以检测到它。启用水印后，词语选择仍然是随机作出的，但随机性的来源有所不同。水印不会使用任意的随机数生成器来挑选下一个词，而是使用密钥以及前面的几个词来确定模型应当选择哪个词。

Take the sentence “The weather today was cold and…”. The next word is very unlikely to be “sugary.” But it is quite likely to be “overcast” or “grey.” Under most circumstances, it doesn’t matter much to the reader which of these latter two words the model ultimately chooses—the meaning of the sentence is largely the same either way. In cases like this, the choice is settled by a random number.

Watermarking uses low-stakes choices like these—which occur many times over a piece of generated text—to leave a pattern in Claude’s responses. That pattern is undetectable to the reader, but is detectable to anyone who has a key that encodes it. When watermarking is used, choices are still made at random, but the source of the randomness is different. Instead of using an arbitrary random number generator to pick the next word, watermarking uses the key and a few words that come before to settle what word the model should pick.