Adaptive Branching Readiness¶

This is the S8 no-submit readiness surface for mid-circuit adaptive branching. It records branch policies and backend prerequisites, with no hardware submission and no adaptive-advantage claim.

Why this file exists¶

This page is used by teams deciding whether to move from simulation to controlled feedback control experiments. It captures the exact branch-policy surface the runtime can execute and the prerequisites that must be present before adaptive control can be promoted.

It is intended for:

control teams testing mid-circuit correction loops;
experiment teams planning conditional gates and reset-dependent branches;
reviewers who need a clear boundary between experimental readiness and promoted hardware claims.

Boundary¶

readiness and branch-planning artifact only; no adaptive-advantage claim and no hardware submission

Branch Policies¶

local_order_threshold: observe partial local Kuramoto order parameter; trigger local_r < target_r - deadband; action corrective_kick.
dla_parity_leakage: observe sector leakage estimate; trigger parity_leakage > max_parity_leakage; action sector_rebalance.
chimera_cluster_detector: observe cluster imbalance with low local order; trigger cluster_imbalance > threshold and local_r below target window; action topology_aware_pulse.

Readiness¶

Ready: False
Required features: ['mid_circuit_measurement', 'conditional_control', 'conditional_reset', 'cross_shot_batches']
Missing features: ['conditional_reset', 'cross_shot_batches']
Branch table rows: 12

Falsifier¶

win-rate <= 50% on preregistered equal-depth open-loop comparison

Prerequisites¶

S1 cross-shot feedback plumbing remains the runner-level dependency
target backend must declare mid-circuit measurement and conditional control
Rust branch-table generation is a performance follow-up, not a correctness prerequisite
hardware execution requires explicit preregistration and approval

Gate¶

Regenerate and compare this readiness artefact with:

scpn-bench s8-adaptive-branching-readiness