On Chromium/linux, pressing pause doesn't pause, instead resetting the animation to it's pre-play state - the current attention highlighting disappears. Pressing play again, restarts at the beginning. Having a commonplace "pause pauses, and play resumes" UI, could allow more time to look over state. A youtube-like slow playback 0.25? option might similarly help. Or perhaps even better, buttons for single stepping. Tnx for your work - nice.
are you worried later-layer attention gets drowned out by earlier layers just because there are more of them contributing to the sum?
Right now only simple correlations are visible.
To me, it's more of a neat visualization, not something that can be used to interpret LLM behavior. Even with a lot of simplification, it can show some interesting patterns.
This is incorrect. The compute required per forward pass to generate each additional token during decode will scales as O(N), even with a KV cache (without a KV cache, it would scale as O(N^2)). Over generating N tokens, it's O(N^2) with the cache (and O(N^3) without).
It's O(N) for a forward pass because that new token still has to "attend to" to each previous token. That requires N dot products: between the cached key vectors and the new query vector for the new position. You also have N reads from memory (K and V) which is probably gonna be your actual bottleneck. (Decode is memory-bound.)
This is why you should avoid long contexts, if you can, even with a warm cache. You will get charged more, in "cache read" tokens.
Or does it accumulate the relations like A relates to B, so also add in B's relations
I mean, there so many headers and layers, it is tricky to make a choice that will resonate with our intuition . Is it some weighted average? Or maybe ablation test?
When I started, I expected I'd have to experiment a lot to find something comprehensible. But this simple computation can already show some patterns.