Either I'm wildly misunderstanding or that can't possibly be true--if you sample at high temperature and it chooses a very-low probability token, it continues consistent with the chosen token, not with the more likely ones
>>bcoate+(OP)
Attention computes a weighted average of all previous latents. So yes, it’s a new token as input to the forward pass, but after it feeds through an attention head it contains a little bit of every previous latent.