The non-stabilizerness, or "magic", of a quantum state is a common metric for the hardness of simulation on a classical computer, and is often quantified by the stabilizer Rényi entropy (SRE). We show how random superpositions of classically simulable states saturate the SRE to the symmetry-restricted Haar-random maximum. For any large-$N$ random pure-state ensemble satisfying the conditions of Schur's lemma, the average SRE of the $m$-component superposition $\langle M_2(m)\rangle=N-\log_2[\langle W(m)\rangle]$ is governed by an average witness $\langle W(m)\rangle$ that universally decomposes into a Haar floor and an ensemble-specific excess. As the number of components $m$ increases, the excess is suppressed, resulting in a logarithmic scaling of the SRE before plateauing at the Haar-valued maximum. We derive a general closed form for $\langle W(m)\rangle$ in the thermodynamic limit, and numerically verify our result for random stabilizer states and random Gaussian states.
AI Summary
The paper asks how stabilizer Rényi entropy (SRE) grows when independently sampled, classically simulable pure states are coherently superposed. It studies ensembles whose mean density matrix is maximally mixed within a symmetry sector, using the Pauli fourth-moment witness $W=\sum_P\langle P\rangle^4$ and the quantity $N-\log_2\mathbb E[W]$.
For component weights $p_i=|\lambda_i|^2$, it derives the large-$N$ relation $\mathbb E[W]=(a-b-c)\sum_i p_i^4+b(\sum_i p_i^2)^2+c$, where $a$ is the single-component witness, $b$ depends on ensemble moments, and $c=4n_{\mathrm{fix}}$ counts symmetry-fixed Pauli strings. Equal weights therefore give $\mathbb E[W]=c+b/m^2+(a-b-c)/m^3$. The resulting SRE grows logarithmically with component number before approaching the symmetry-restricted Haar value $N-2-\log_2 n_{\mathrm{fix}}$.
For random stabilizer components, $\mathbb E[W]=4+(2^N-4)/m^3+O(2^{-N})$, giving initial growth $3\log_2m$ and a crossover near $m=2^{(N-2)/3}$. For fixed-parity fermionic Gaussian components, the floor is $8$, the single-component witness scales as $\sqrt{\pi}N^{3/2}/4$, and both inverse-square and inverse-cube contributions occur. Computational-basis superpositions instead exhibit $2\log_2m$ growth.
The derivation expands fourth moments into index-contraction diagrams, applies ensemble moment identities, and drops thermodynamically vanishing terms; numerical tests cover stabilizer and Gaussian ensembles. Results are asymptotic. The defined entropy uses the logarithm of the mean witness, rather than the mean entropy. Superpositions are normalized through their weights rather than realization by realization, with a stated average witness error of order $2^{-N}$.
Little sign of AI-generated textOverclaimsPossible error
Central claim may not hold: The general derivation incorrectly infers zero means and a common directed second moment for filtered Paulis from sector isotropy alone.
The printed filtered-sum formula for b gives b tending to -6 for random stabilizer states, rather than the zero required by the stabilizer specialization.
The reported 'average SRE' is an annealed quantity rather than the ensemble average of the state SRE, and their equivalence is not established for the claimed class of ensembles.
An isotropic sector with a fixed non-Pauli single-qubit state has a sector-Haar witness tending to 6 despite n_fix=1, directly contradicting the claimed universal floor of 4 n_fix.
The claimed N=8 saturation by m=4 contradicts the paper's own formula, which places the SRE almost one bit below its stated ceiling.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
The non-stabilizerness, or "magic", of a quantum state is a common metric for the hardness of simulation on a classical computer, and is often quantified by the stabilizer Rényi entropy (SRE). We show how random superpositions of classically simulable states saturate the SRE to the symmetry-restricted Haar-random maximum. For any large-N random pure-state ensemble satisfying the conditions of Schur's lemma, the average SRE of the m-component superposition ⟨M2(m)⟩=N−log2[⟨W(m)⟩] is governed by an average witness ⟨W(m)⟩ that universally decomposes into a Haar floor and an ensemble-specific excess. As the number of components m increases, the excess is suppressed, resulting in a logarithmic scaling of the SRE before plateauing at the Haar-valued maximum. We derive a general closed form for ⟨W(m)⟩ in the thermodynamic limit, and numerically verify our result for random stabilizer states and random Gaussian states.
The paper asks how stabilizer Rényi entropy (SRE) grows when independently sampled, classically simulable pure states are coherently superposed. It studies ensembles whose mean density matrix is maximally mixed within a symmetry sector, using the Pauli fourth-moment witness W=∑P⟨P⟩4 and the quantity N−log2E[W].
For component weights pi=∣λi∣2, it derives the large-N relation E[W]=(a−b−c)∑ipi4+b(∑ipi2)2+c, where a is the single-component witness, b depends on ensemble moments, and c=4nfix counts symmetry-fixed Pauli strings. Equal weights therefore give E[W]=c+b/m2+(a−b−c)/m3. The resulting SRE grows logarithmically with component number before approaching the symmetry-restricted Haar value N−2−log2nfix.
For random stabilizer components, E[W]=4+(2N−4)/m3+O(2−N), giving initial growth 3log2m and a crossover near m=2(N−2)/3. For fixed-parity fermionic Gaussian components, the floor is 8, the single-component witness scales as πN3/2/4, and both inverse-square and inverse-cube contributions occur. Computational-basis superpositions instead exhibit 2log2m growth.
The derivation expands fourth moments into index-contraction diagrams, applies ensemble moment identities, and drops thermodynamically vanishing terms; numerical tests cover stabilizer and Gaussian ensembles. Results are asymptotic. The defined entropy uses the logarithm of the mean witness, rather than the mean entropy. Superpositions are normalized through their weights rather than realization by realization, with a stated average witness error of order 2−N.
We characterize the magic of random fermionic Gaussian states through their Pauli spectrum and their stabilizer entropies. For the Haar ensemble of Majorana Gaussian states we derive closed finite-N expressions for the average stabilizer purities and reconstruct the Pauli spectrum exactly, as a mixture of products of beta-distributed variables resolved by Majorana weight. The stabilizer entropies, and their filtered versions, freeze at Rényi index qc=2: for q>2 the magic density is 1/(q−1), controlled by rare weight-two Majorana strings. Resolving these contributions requires exponentially many characteristic samples, making direct sampling estimates of the stabilizer entropies inefficient in this frozen regime. For Haar-random Gaussian states at fixed filling, we express the averaged stabilizer purities at positive integer q as coefficient integrals over SU(2q), with dimension independent of N, and evaluate the q=2 case exactly. We establish that the frozen density describes typical states in the Majorana and half-filled number-conserving ensembles and relate these ensembles to the corresponding SYK2 ground states. Finally, we show that typical fermionic Gaussian states retain certified magic after discarding any fixed fraction κ<κ⋆G≃0.7613 of their modes in the thermodynamic limit, remaining magical even beyond the 2/3 threshold for Haar-random states in the full Hilbert space.
Nonstabilizerness or `magic' is a key resource for quantum computing and a necessary condition for quantum advantage. Non-Clifford operations turn stabilizer states into resourceful states, where the amount of nonstabilizerness is quantified by resource measures such as stabilizer Rényi entropies (SREs). Here, we show that SREs saturate their maximum value at a critical number of non-Clifford operations. Close to the critical point SREs show universal behavior. Remarkably, the derivative of the SRE crosses at the same point independent of the number of qubits and can be rescaled onto a single curve. We find that the critical point depends non-trivially on Rényi index α . For random Clifford circuits doped with T-gates, the critical T-gate density scales independently of α . In contrast, for random Hamiltonian evolution, the critical time scales linearly with qubit number for α > 1 , while it is a constant for α < 1 . This highlights that α -SREs reveal fundamentally different aspects of nonstabilizerness depending on α : α -SREs with α < 1 relate to Clifford simulation complexity, while α > 1 probe the distance to the closest stabilizer state and approximate state certification cost via Pauli measurements. As technical contributions, we observe that the Pauli spectrum of random evolution can be approximated by two highly concentrated peaks which allows us to compute its SRE. Further, we introduce a class of random evolution that can be expressed as random Clifford circuits and rotations, where we provide its exact SRE. Our results opens up new approaches to characterize the complexity of quantum systems.
We investigate universal entanglement properties inherent to superpositions of randomized states. We find that an m-fold superposition of typical states may be classified into two distinct entanglement classes via the 2nd Rényi entropy density s2. The maximally entangled regime is defined by s2∼ln(2), for which superposition adds no additional entanglement. The sub-maximally entangled regime, s2<ln2, instead constrains the reduced density matrices of independent components to be orthogonal in the thermodynamic limit, which fixes the entanglement of the superposition to a logarithmic enhancement ΔS(m)=ln(m). As a consequence, an exponentially large number of superpositions is required to transition from the sub-maximally entangled class to maximal entanglement. We explicitly calculate s2 and the logarithmic enhancement, and demonstrate orthogonality for two canonical examples of the sub-maximally entangled regime (superpositions of pure Gaussian states and of random matrix-product states). We also examine the entanglement of superpositions of random stabilizer states, and discuss their relaxation to the Haar limit.
Magic is a property of a quantum state that characterizes its deviation from a stabilizer state, serving as a useful resource for achieving universal quantum computation e.g., within schemes that use Clifford operations. In this work, we study magic, as quantified by the stabilizer Renyi entropy, in a class of models known as generalized Rokhsar-Kivelson systems, i.e., Hamiltonians that allow a stochastic matrix form (SMF) decomposition. The ground state wavefunctions of these systems can be written explicitly throughout their phase diagram, and their properties can be related to associated classical statistical mechanics problems, thereby allowing powerful analytical and numerical approaches that are not usually available in conventional quantum many body settings. As a result, we are able to express the SRE in terms of wave function coefficients that can be understood as a free energy difference of related classical problems. We apply this insight to a range of quantum many body SMF Hamiltonians, which affords us to study numerically the SRE of large high-dimensional systems, and in some cases to obtain analytical results. We observe that the behaviour of the SRE is relatively featureless across quantum phase transitions in these systems, although it is indeed singular (in its first or higher order derivative, depending on the nature of the transition). On the contrary, we find that the maximum of the SRE generically occurs at a cusp away from the quantum critical point, where the derivative suddenly changes sign. Furthermore, we compare the SRE and the logarithm of overlaps with specific stabilizer states, asymptotically realised in the ground state phase diagrams of these systems. We find that they display strikingly similar behaviors, which in turn establish rigorous bounds on the min-relative entropy of magic.
Magic states are essential resources for fault-tolerant quantum computation. We introduce the amortized stabilizer Rényi entropy (SRE), which quantifies the largest SRE increase generated by a unitary, including arbitrary pure inputs and reference systems. Initial nonstabilizerness can enhance this increase, whereas analogous amortized logarithmic robustness and extent collapse to stabilizer-input optimizations. We establish structural properties and exact amortized 2-SRE values for the T and CCZ gates. We then derive explicit lower bounds on approximate T-state cost, with diamond-distance error, allowing stabilizer ancillary states, Pauli measurements, and classical feedforward. A Choi-state witness gives computable bounds for quantum Fourier transforms, controlled rotations, fractional controlled-SWAP gates, and finite Heisenberg-chain evolutions. We distinguish the unrestricted optimization from its Choi witness and specify the access and precision requirements for estimating Pauli moments. An odd-moment variant yields a confidence-qualified cost witness from ordinary copies.
Non-stabilizerness (colloquially "magic") characterizes genuinely quantum (beyond-Clifford) operations necessary for preparation of quantum states, and can be measured by stabilizer Rényi entropy (SRE). For permutationally symmetric states, we show that the SRE depends, for sufficiently large systems, only on a constant number of expectation values of collective spin operators. This compact description is leveraged for analysis of spin-squeezing protocols, which inherently generate non-stabilizerness. Under one-axis twisting (OAT), the generation of optimal squeezing is accompanied by a logarithmic divergence of SRE with system system size. Continued time evolution under OAT produces metrologically useful "kitten" states-superpositions of rotated GHZ states-that feature many-body Bell correlations but exhibit a smaller, system-size-independent SRE that decreases with increasing Bell-correlation strength. Our results reveal connections between non-stabilizerness, multipartite correlations, and quantum metrology, and provide a practical route to quantify non-stabilizerness in experiments for precision sensing.