The Setting
Uniform circuit localization bounds the passage from a product projection’s zero-mismatch sector to a high-mismatch sector through a conjugated product projection. The mismatch cutoff and the decay rate have separate exponents. The finite-scale bootstrap concerns assertions indexed by these two exponents.
A classical bit is an element of ✦, and a bit string of length assigns a classical bit to each of wire positions.✦ A quantum operator on total qubits is a complex matrix whose rows and columns are indexed by bit strings of length .✦ A single-qubit operator is a complex matrix indexed by classical bits.✦
For a one-qubit vector , the outer product has entry in row and column .✦ The tensor product of single-qubit operators is the operator whose entry from a bit string to a bit string is .✦
Product Projections and Circuit Scales
Product projection data on qubits consist of a support and local vectors satisfying (1) on the support: (1)✦ For product projection data on qubits and a set of wire positions, the mismatch pattern projection is the tensor product in (2): (2) For product projection data on qubits and a real cutoff , the zero-mismatch product projection and the high spectral projection are the operators defined by (3); the high cutoff includes equality: (3)
For a natural depth and total qubits, the enlarged reflection-circuit class is the recursively defined class of operators of depth at most in (4). Here denotes the matrix of a layer of single-qubit unitaries and the matrix of a layer of reflections about product projections with pairwise disjoint supports; identity layers are allowed, and the rightmost matrix acts first. $$\begin{aligned} U_0(N)&=\{L:L\text{ is a local-layer matrix}\},\\ U_{d+1}(N)&=\{VRL:V\in U_d(N),\ R\text{ is a reflection-layer matrix},\\ &\hspace{42mm}L\text{ is a local-layer matrix}\}. \end{aligned}$$(4)✦
For quantum operators and on qubits, operator conjugation of by is the operator , where the dagger denotes the adjoint.✦ For a natural depth and real exponents , uniform localization is the assertion that (5) holds for all sufficiently large , simultaneously for every and all product projection data on qubits: (5)✦ The norm is the Euclidean operator norm.
The Claim
Theorem 1 (Finite-scale bootstrap). Let be an assertion for pairs of real numbers. Suppose holds whenever and . Suppose also that for every choice of real numbers satisfying and , validity of implies . Then (6)✦
Theorem 2 (Localization above one). For every natural depth and all real exponents with , (7)✦
What Is Checked
The finite-scale bootstrap (6) is stated over arbitrary assertions on pairs of real numbers with the large-threshold and backward-step hypotheses, while (7) concerns uniform localization over the enlarged reflection-circuit classes and all normalized product projection data.