Comment by ahartmetz
3 days ago
I'm not sure if that is what she meant. In any case, it does sound very useful for code. The input becomes messy to represent as text, but prompts don't usually make it into version control, so not much is lost.
3 days ago
I'm not sure if that is what she meant. In any case, it does sound very useful for code. The input becomes messy to represent as text, but prompts don't usually make it into version control, so not much is lost.
I'd bet this is what she meant because I've done professional translation in the past.
The input and output of a GPT it not a sequence of messages in turns: fundamentally the input is the entire context (both from the user and the model), and the output is a probability distribution for 1 single next token.
The chat is only one of many interaction patterns possible with that primitive: one where the "most likely" output token is concatenated to the previous input and this is repeated until a limit is reached or the model produces a special "end of sequence" token.
Representing a more complete interaction pattern as a chat is impossible because the chat abstraction is lossy and could never represent it. Sometimes the abstraction leaks, as with the concept of "temperature", that's not a parameter of the model but of the chat.