lukemelas / do-you-even-need-attention Goto Github PK

Hi Luke, thank you for sharing this amazing work.

In your arxiv document, I cannot find any mention of positional encoding, but I see that you use them in your code. Did you conduct any ablation study on the PE? i.e., how much does it affect the performance, with and without?

Thank you in advance.

In the experiment section of the report, you mention:

Such a comparison is not exactly fair because the feed-forward model uses stronger training augmentations.

I wonder what the augmentations are, for details about the augmentation seems missing from the paper.

Hi, I was going through your exp report. You have made a point that since you are able to get a good performance without using attention layer so good performance of ViT may be more to do with it's embedding layer than attention .

But I believe, It's also may be to do with how you have established an interaction between patches through a transpose very similar to what was done in MLP-Mixer .

Would love to know your thoughts on this ?