3 ms·

So instead of generating the next token from its own previous predictions (which is what it would do in real life), the code they used for the evaluation actual

So instead of generating the next token from its own previous predictions (which is what it would do in real life), the code they used for the evaluation actually predicts from the ground truth?

ghayes 3y ago
Which would basically turn the model into a plainly normal LLM without any need for utilizing the brainwave inputs, right?