alt Hacker News>This suggests that deeper isn't always better for looped transformers, which leaves me less worried about a race to the bottom toward looped transformers with hundreds of recurrent loops.
I disagree with this. Deeper will always be at least as good because the extra loops can exit early or just no-op. Any performance degradation they're seeing at higher loop counts today is merely training stability issues, which can be overcome.
Deeper almost certainly is better, and we will probably see not just hundreds but millions of recurrent loops in the future.
Deeper independent, sure. Deeper shared though? Information and signal theory still apply here. At infinite cycles, without new input, you'll end up with a locked state or oscillations. Some point before that, any "attractors" in the latent space, with slightly higher statistics, will pull things towards a space that might eventually be only loosely related to the goal, because each loop would be lossy, right?