← All research questions
Q36 / Algorithms & primitives / question

Can local spin tests perform a spectral measurement?

Can pairwise STP approximate a spectral threshold measurement for an exchange Hamiltonian? Bound the error and resource cost, and ask separately for a protocol that retains the filtered state.

Discuss this question
Working mathematical formulation

For a physically specified exchange Hamiltonian H=∑eaePseH=\sum_e a_eP_s^e with polynomial input description and ∥H∥≤1\lVert H\rVert\le 1, let 0≤f≤10\le f\le 1 approximate a threshold with spectral margin Δ\Delta. Construct an STP acceptance effect EE with ∥E−f(H)∥≤ε,\lVert E-f(H)\rVert\le\varepsilon, or prove an obstruction in the stated model. Bound qubits, tests, rounds, and success probability in n,1/Δ,1/εn,1/\Delta,1/\varepsilon. Separately approximate the full spectral-filtering measurement, including the output state and its correlations with a reference.

Why this is interesting

A sequence of STP measurements defines an acceptance effect. Comparing that effect with a function of the Hamiltonian tests whether the protocol performs spectral filtering.

A place to start

Compute the effect of alternating two overlapping projectors exactly, then compare it with filters of their sum. A local-term average estimates mean energy; it is not a measurement of the Hamiltonian’s spectrum.

Related work

These works provide context and established results. They need not be the origin of this exact formulation.

Discuss this question

Reading is public. Posting requires a GitHub account. You can also write to Kunal.