[ottaborra]

RNN with extra steps?

[tripplyons]

There are many papers that use a recurrence across sub-sequences and attention within sub-sequences. Google did this with Infini-Attention and one of the variants from the Titans paper. However, I think the earliest example of this is Transformer-XL.

[biofox]

Isn't that all of modern AI?

[immibis]

Transformers are completely unlike RNNs.

[tripplyons]

There are some interesting connections between them. If you remove the softmax from the attention formula, you end up with linear attention, which has a recurrent form.

I haven't read it, but the Mamba 2 paper claims to establish a stronger connection.

[kadushka]

* If you remove the softmax from the attention formula, you end up with linear attention*

Sorry, what?

[tripplyons]

Here is a paper explaining it: https://arxiv.org/abs/2006.16236