5 ms·
But the point is to focus on the intrinsic dimension[1], not dimensions of the vector itself. I meant dense in the sense that the two are close, relative to ano
by magicalhippo 2y ago
But the point is to focus on the intrinsic dimension[1], not dimensions of the vector itself. I meant dense in the sense that the two are close, relative to another vector where they are not so close. Perhaps a poor choice of words on my part.
[1]: https://en.wikipedia.org/wiki/Intrinsic_dimension https://en.wikipedia.org/wiki/Intrinsic_dimension
- slashdave 2y agoAh, I see. Well, the data has an intrinsic dimension of a specific size. You don't get to choose that. And, in any case, you want something quite a bit larger than the intrinsic dimension, because deep-learning needs redundancy in its weights in order to train correctly.
- magicalhippo 2y agoRight, but part of the argument in the paper, as I understand it, is that the self-attention layers can increase the intrinsic dimension of the input data if you feed it additional, relevant context. I guess you could also use this result to find that a smaller network might be sufficient for your particular problem.
- slashdave 2y agoIf you have additional context that is relevant, feed it to the network. Why wouldn't you? As to the size of the network, this is not a simple benefit, because you need to account for the trade off between model size and training efficiency.