← Lab
blindband
confirmedshipPython · PyTorchModel behavior & reasoningJuly 22, 2026
the question
Does StreamingLLM attention with a full retained cache have an exact machine-zero blind band that no depth can close?
what came out
Influence support is exactly [0,S-1] union [i-(w-1)L,i], with a machine-zero blind band between them measured 0.00e+00 at depths 2,4,8,12. The band grows linearly in length and shrinks only linearly in depth. One global-attention or recurrent layer closes it, the exact capacity reason hybrids interleave.
method & receipts
- Result: confirmed
- Tested — a correctness/benchmark suite ships alongside the code.
→ read the code and re-run it
github.com/v-code01/blindband