Core Concepts
Any pseudoentangled state ensemble with a gap of t bits of entropy requires Ω(t) non-Clifford gates to prepare.
Abstract
The content discusses the relationship between pseudoentanglement and non-Clifford complexity in quantum computing. The key highlights are:
The authors present an algorithm that can efficiently estimate the entanglement entropy across any bipartition of qubits for quantum states with bounded stabilizer dimension. Specifically, if a state |ψ⟩ has stabilizer dimension at least n-k, the algorithm outputs upper and lower bounds on the entanglement entropy that are within k bits of the true value.
As a corollary, the authors show that any family of Clifford circuits that produces a pseudoentangled ensemble {|Ψk⟩, |Φk⟩}k with entropy gap f(n) vs. g(n) satisfying f(n) - g(n) ≥ t must use Ω(t) auxiliary non-Clifford single-qubit gates.
This matches the lower bound on non-Clifford gates needed for pseudorandom states, showing that pseudoentanglement is as computationally expensive as pseudorandomness in terms of non-Clifford complexity.
The authors also discuss a concurrent work by Gu, Oliveiro, and Leone that obtains related results on the relationship between magic and entanglement.
Stats
Any n-qubit state |ψ⟩ that is stabilized by at least 2^(n-t) Pauli operators requires Ω(t) non-Clifford gates to prepare.
Quotes
"Any pseudoentangled state ensemble with a gap of t bits of entropy requires Ω(t) non-Clifford gates to prepare."
"Pseudoentangled states can be instantiated in polynomial time and logarithmic depth, assuming the existence of quantum-secure one-way functions."