overdisp

mixedshipPythonInference & servingJuly 12, 2026

the question

Is the accept/reject stream in greedy speculative decoding autocorrelated and over-dispersed, and does that raise the optimal draft length above the i.i.d. prediction?

what came out

The greedy accept/reject stream is autocorrelated (pooled p11-p10 +0.155) and over-dispersed (run-length D/D_geom up to 24x), confirming it is not i.i.d. But the pre-registered consequence reverses: the optimal draft length moves down, not up. The marginal-alpha i.i.d. model over-predicts per-block yield (E_true/E_iid 0.883 at length 16).

method & receipts

Result: mixed
Pre-registered — the prediction was written down and committed before the run.
Reproducible — one script re-runs the whole thing from scratch.
Tested — a correctness/benchmark suite ships alongside the code.

→ read the code and re-run it

github.com/v-code01/overdisp