[AdieuToLogic]

>> A popular LLM joke at the moment is "How many Rs in strawberry?"

> ChatGPT answer: There are two "R"s in the word "strawberry."

Given enough instances of the "LLM joke" in a training data set, the joke itself having a consistent form (sequence of tokens) and likely followed by the answer having a similarly consistent form (sequence of tokens), the probability of the latter being produced as quoted is high.

[imtringued]

There is a lot of broken English on the internet and yet LLMs are better at English than the average native speaker. This failure mode has nothing to do with the training data.