Pith. sign in

REVIEW 2 major objections 3 minor 68 references

A single chaotic Hamiltonian plus simple pulses can generate random-matrix-level unitary designs

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · deepseek-v4-flash

2026-08-03 01:50 UTC pith:XLSSDBEH

load-bearing objection Solid, honestly-scoped theory paper: Eq. (7) checks out internally, but the nonresonance assumption is a true scope limit, not a minor footnote. the 2 major comments →

arxiv 2607.24851 v2 pith:XLSSDBEH submitted 2026-07-25 quant-ph

Unitary designs from perturbed time evolutions of a chaotic Hamiltonian

Zhongyi Yang , Biao Wu This is my paper
classification quant-ph MSC 81P4581Q5015B52 PACS 03.65.Aa03.65.Yz05.45.Mt
keywords unitary designsframe potentialchaotic Hamiltoniantemporal ensemblenonresonance conditionPauli ensembleClifford ensemblerandom unitaries
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper claims that unitary k-designs, the workhorse substitute for fully random unitaries in quantum information, can be generated using just one fixed chaotic Hamiltonian together with randomly sampled evolution times and intermediate unitary pulses. The central result is a formula expressing the frame potential of this single-Hamiltonian ensemble in terms of the frame potentials of the pulse ensemble. If true, this sharply reduces experimental control requirements: the pulse ensemble only needs to suppress its frame-potential growth relative to the Hilbert-space dimension, not itself be random or a design. The paper further claims that, at the level of frame potentials, multi-Hamiltonian protocols can be recursively replaced by one Hamiltonian interleaved with fixed traceless perturbations. For platforms like trapped ions and cold atoms where switching Hamiltonians is costly, this offers a minimal route to random-unitary generation.

Core claim

For a fixed chaotic Hamiltonian H whose spectrum satisfies the k-th order additive nonresonance condition, and an intermediate unitary ensemble E_int satisfying a trace-suppression condition, the eigenbasis-averaged k-th frame potential of the one-Hamiltonian temporal ensemble {e^{-iHt2} U e^{-iHt1}} equals k! + sum_{ell=1}^{k} (k!)^2 F^{(ell)}_{E_int} / ((k-ell)! D^{2ell}) + o(1) in the perfect time-filter and large-dimension limits. Since the Haar frame potential is k!, the ensemble forms an approximate unitary k-design whenever the intermediate ensemble's frame potentials grow slower than D^{2ell}. The paper also establishes a replacement rule: a second independent chaotic Hamiltonian evo

What carries the argument

The k-th frame potential F_E^{(k)} = E_{U,V} |Tr(U^dagger V)|^{2k}, whose Haar minimum k! certifies approximate unitary k-design. The proof uses Haar twirling with Weingarten calculus (leading-order D^{-q} scaling), the k-th order additive nonresonance condition to force the time-averaged energy-conservation constraints into permutation sectors, and the trace-suppression condition to show only pair-matching permutation cycles survive. The final summation over permutations reduces to a binomial identity, giving the closed-form relation Eq. (7).

Load-bearing premise

The entire derivation collapses if the fixed Hamiltonian's spectrum contains an accidental energy resonance sum_{a_r}-sum_{b_r}=0 for non-permuted replica strings; such degeneracies would break the permutation-sector reduction that produces Eq. (7).

What would settle it

Compute the k-th frame potential of the one-Hamiltonian ensemble E_N(H, P_n^*) for a specifically engineered Hamiltonian with a known exact resonance (e.g., equally spaced energy levels) and check whether the deviation from k! remains small as D grows; Eq. (7) predicts failure when the nonresonance condition is violated.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • If Eq. (7) is correct, approximate unitary k-designs can be produced with a single fixed Hamiltonian, requiring no Hamiltonian switching, only time sampling and intermediate pulses such as non-identity Pauli operators or Clifford operations.
  • The criterion for the intermediate ensemble is directly testable: it must satisfy F_E^{(ell)} = O(D^{2ell - epsilon}) for all ell <= k; nontrivial Pauli and Clifford ensembles pass, while identity-only or fixed-size local Pauli ensembles fail.
  • Multi-Hamiltonian protocols (three-Hamiltonian, Pauli-assisted two-Hamiltonian) can be replaced, at the frame-potential level, by one-Hamiltonian protocols interleaved with fixed traceless unitaries, reducing control complexity on NISQ devices.
  • Finite-time numerical results for both GUE and Rydberg Hamiltonians show that the design-forming effect appears for moderate time windows, not only in the ideal T -> infinity limit.
  • The replacement rule suggests that intermediate pulses act as catalysts: a fixed Pauli kick redirects chaotic evolution into distinct histories whose overlaps, as measured by the frame potential, approach the Haar level.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The result suggests a general design principle: any physically implementable pulse ensemble whose cardinality grows polynomially with D, or that forms a group with bounded trace moments, can serve as the intermediate randomness source; this is broader than the Pauli and Clifford examples listed.
  • If the frame-potential replacement rule holds beyond the k-th order to all moments, it would imply a kind of simulation equivalence between independent Hamiltonian controls and fixed perturbations, potentially simplifying randomized benchmarking and shadow-estimation protocols on analog simulators.
  • A natural testable extension is to replace the fixed traceless unitary with a fixed trace-zero channel or a small set of pulses, checking whether the frame-potential reduction in Eq. (15) is preserved.
  • The additive nonresonance condition is the key spectral assumption; for realistic many-body systems with symmetries or near-degeneracies, designing slightly inhomogeneous Hamiltonians (as done for the Rydberg model) may be necessary, and quantifying the failure mode when resonances exist remains open.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

2 major / 3 minor

Summary. The paper proposes a protocol for generating approximate unitary k-designs from a single fixed chaotic Hamiltonian H by interleaving two time evolutions e^{-iHt2} U e^{-iHt1} with an intermediate unitary U drawn from an ensemble E_int. The central result (Theorem 1, Eq. (7)) states that, in the T→∞ perfect time-filter limit and under an additive nonresonance condition on the spectrum (Eq. (5)) and a trace-suppression condition on E_int (Eq. (4)), the Haar-eigenbasis-averaged k-th frame potential equals k! + Σ_{ℓ=1}^k (k!)^2 F^{(ℓ)}_{E_int}/((k-ℓ)! D^{2ℓ}) + o(1). This yields a sufficient condition (Eq. (8)): if F^{(ℓ)}_{E_int}=O(D^{2ℓ-ε}), the protocol forms an approximate design. Pauli and Clifford intermediates satisfy it; identity and fixed-size local ensembles do not. The paper also claims a recursive replacement rule (Eqs. (16)-(18)) equating multi-Hamiltonian temporal protocols to one-Hamiltonian protocols with fixed traceless perturbations. Numerical GUE and Rydberg simulations support the formula and finite-time convergence.

Significance. If correct, the result is significant: it reduces Hamiltonian control from three independent Hamiltonians to one fixed Hamiltonian plus intermediate pulses, and gives a parameter-free closed-form frame potential. The derivation is structurally coherent: the time-filter reduction, Weingarten leading-order selection, and direct/crossed matching analysis reproduce Eq. (7), and the formula is benchmarked against the Haar value k! and against independent temporal-ensemble results. The criterion cleanly separates design-forming intermediates (Pauli, Clifford) from design-limiting ones (identity, local Pauli). However, the advertised scope for fixed physical Hamiltonians and the recursive replacement rule both require additional justification; the proof's rigor under the stated assumptions is not in question.

major comments (2)
  1. [Def. 2, Eq. (5); Theorem 1; Appendix A; Appendix B] The additive nonresonance condition is load-bearing: Theorem S1 uses it to reduce the T→∞ time-filter average to the permutation sector. If a realistic spectrum has degeneracies or accidental many-body resonances, non-permutation strings survive and Eq. (7) has no error term, so the claimed approach to k! is not established. The paper provides no proof that a fixed local chaotic Hamiltonian (e.g., the Rydberg model, Eq. (B1)) satisfies Eq. (5); Appendix B adds random inhomogeneities precisely 'to break spatial symmetries and suppress accidental many-body resonances', and robustness to symmetries/near-degeneracy is deferred. Appendix A's typicality argument averages over Haar-random eigenbases for a fixed spectrum (Markov on Δ_W); it does not certify a particular physical eigenbasis. Thus the central claim—a single fixed chaotic Hamiltonian generates a design—is proven only under an unver
  2. [Eqs. (16)-(18)] The recursive replacement E_L(H1,...,Hm) ≈ E^{(m)}_N(H1,U0) is not a direct consequence of Eq. (16). Eq. (16) equates the k-th frame potentials of the two ensembles E_L(H1,H2) and E_N(H1,{U0}), but frame potentials are not compositional under independent left/right multiplication by random unitaries. Therefore 'iterating Eq. (16)' requires an argument that the replacement preserves the frame potential of the larger product ensemble. No such proof is given in the main text or SM; Fig. 3(c) provides numerical evidence for m=2,3 only. Since the recursive construction is advertised in the abstract and summary, this gap should be closed or the claim should be explicitly stated as a numerically supported conjecture.
minor comments (3)
  1. [Fig. 2 caption] The caption states that black dashed lines show the analytical prediction from Eqs. (7) and (13), but the rendered figure does not clearly display them. Please check that the curves are visible in the final production version.
  2. [Appendix C] The finite-T analysis is entirely numerical. It is fine as evidence, but the text should more carefully distinguish the proven T→∞ statement from the numerically observed finite-T behavior, especially because the abstract emphasizes experimental feasibility.
  3. [Notation, Eq. (3) vs Eq. (6)] The distinction between F^{(k)}_{E_N}(H,E_int) and the eigenbasis-averaged quantity with an overline is introduced but could be made more explicit at first use, to avoid confusion in the statement of Theorem 1.

Circularity Check

0 steps flagged

No significant circularity: Theorem 1 is a derived frame-potential transfer relation, benchmarked against the external Haar value, with no fitted parameters or load-bearing self-citations.

full rationale

The central claim, Eq. (7), is a theorem (proved in SM Sec. S3) expressing the k-th frame potential of the one-Hamiltonian ensemble E_N in terms of the frame potentials of the intermediate ensemble E_int. The proof uses the additive nonresonance condition (Def. 2), the trace-suppression condition (Def. 1), Haar twirling/Weingarten calculus, and combinatorial bounds (Propositions S1–S4). No parameter is fitted to the target Haar value k!, and the Haar value itself is imported from external references [16,33]. The suppression criterion (Def. 3) is a sufficient condition derived from Eq. (7), not an assumption equivalent to the conclusion. The replacement rule Eq. (16) is obtained by substituting the fixed-traceless-perturbation frame potential into Eq. (7) and comparing with the independent two-Hamiltonian result of Ref. [27]; it is numerically verified in Fig. 3, so it is not a renaming. The self-citations present (Refs. [35], [40], [44]) are contextual or application-oriented and are not load-bearing in the derivation. The paper explicitly flags the nontriviality of the nonresonance assumption for concrete Hamiltonians and the need for robustness analysis; this is a scope limitation, not circularity. The GUE and Rydberg numerics confirm the formula without fitting. Overall, the derivation is self-contained and non-circular.

Axiom & Free-Parameter Ledger

3 free parameters · 8 axioms · 1 invented entities

The theorem itself is parameter-free: no constants are fitted. The free parameters listed are simulation/model choices for the numerical demonstrations, of which the Rydberg disorder ranges are the most load-bearing because they enforce the nonresonance hypothesis. The axioms are dominated by two domain assumptions — additive nonresonance and trace-suppression — both stated explicitly in the main text. No invented entities appear.

free parameters (3)
  • Rydberg Hamiltonian parameters = Omega0 = Delta0 = 1, V0 = 1.5
    Physical model parameters in Appendix B; not fitted to the design claim, but chosen without a stated principle; they affect only the numerical demonstration, not the theorem.
  • Rydberg disorder ranges for inhomogeneities = epsilon in [-0.12, 0.12], eta in [-0.35, 0.35]
    Hand-chosen 'to break spatial symmetries and suppress accidental many-body resonances' (Appendix B); this is precisely what enforces the additive nonresonance hypothesis for a physical Hamiltonian.
  • Time sampling window T = 10^3 to 10^10 depending on figure
    Chosen large to approximate the perfect time-filter limit; the central theorem is T->infinity, and finite-T behavior is addressed only numerically (Appendix C).
axioms (8)
  • domain assumption k-th order additive nonresonance condition (Def. 2, Eq. 5)
    Required for the perfect-time-filter average to reduce to permutation-matched terms (Theorem S1, SM Eqs. S5-S10). Valid a.s. for GUE-type spectra but not guaranteed for arbitrary many-body Hamiltonians; the Rydberg demonstration adds disorder to enforce it.
  • domain assumption Trace-suppression condition (Def. 1, Eq. 4)
    Required for Prop. S4 to suppress non-pairing permutations in the Weingarten expansion; satisfied by the Pauli and Clifford examples.
  • domain assumption Perfect time-filter limit T -> infinity
    Theorem 1 is proven in this limit; the finite-T convergence claims in Appendix C are numerical only.
  • domain assumption Haar-eigenbasis averaging with typicality transfer (Theorem 1, Appendix A)
    The theorem is literally about the eigenbasis-averaged frame potential; the physical claim for a fixed Hamiltonian rests on the Appendix A typicality argument, which uses the universal lower bound F >= k! plus Markov's inequality.
  • standard math Weingarten calculus / Schur-Weyl duality for Haar twirling (SM Sec. S2)
    Standard tool; uses the leading-order Weingarten scaling stated in Eq. (S14).
  • standard math Frame potential lower bound F^(k) >= k! for any unitary ensemble
    Attributed to Refs. [16,33]; used in the typicality argument and as the design criterion.
  • standard math Clifford group frame-potential formula (Eq. 11), cited from Ref. [39]
    Used to show the Clifford ensemble satisfies the suppression criterion; external cited result, not re-derived.
  • standard math Reference frame potentials for E_L(H1,H2) and E_L(H1,H2,H3) from Refs. [27,29]
    Used for the equivalence claims (Eqs. 12, 15-17); the paper matches its computation to these values and verifies numerically in Fig. 3.
invented entities (1)
  • None no independent evidence
    purpose: No new physical entities, forces, or dimensions are introduced.
    The fixed traceless perturbation U0 is a protocol element (a Pauli gate), not an invented entity. The 'process-tensor butterfly space' and 'perturbations as catalysts' language (Summary) are interpretative analogies without independent evidential burden.

pith-pipeline@v1.3.0-alltime-deepseek · 20181 in / 38203 out tokens · 389485 ms · 2026-08-03T01:50:29.837967+00:00 · methodology

0 comments
read the original abstract

Unitary designs provide efficient substitutes for Haar-random unitaries in quantum information processing, randomized measurements, and many-body quantum dynamics. We propose a protocol based on time evolutions of a single chaotic Hamiltonian with intermediate unitary perturbations to generate unitary designs. We derive the frame potential of the resulting ensemble in terms of those of the intermediate ensemble. The intermediate ensemble need not itself be Haar random or even approximately form a unitary design; it is sufficient that its frame-potential growth remains well suppressed relative to the maximal possible scaling. The nontrivial Pauli set and the Clifford ensemble are simple examples satisfying this criterion, whereas ensembles whose cardinality remains independent of the Hilbert-space dimension generally fail. We further show that, at the level of frame potentials, multi-Hamiltonian temporal protocols can be recursively replaced by one-Hamiltonian protocols interleaved with fixed trace-suppressed perturbations.

Figures

Figures reproduced from arXiv: 2607.24851 by Biao Wu, Zhongyi Yang.

Figure 1
Figure 1. Figure 1: FIG. 1. Schematic of the three Hamiltonian-based uni [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Frame potential [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Equivalent constructions of unitary ensembles generated by chaotic Hamiltonian dynamics. The evolution [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Frame potential [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5. The normalized deviation of the frame potential [PITH_FULL_IMAGE:figures/full_fig_p009_5.png] view at source ↗

Review history (2 revisions) →

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

68 extracted references · 5 linked inside Pith

  1. [1]

    D. A. Roberts and D. Stanford, Diagnosing chaos using four-point functions in two-dimensional con- formal field theory, Phys. Rev. Lett.115, 131603 (2015)

    2015

  2. [2]

    Cotler, N

    J. Cotler, N. Hunter-Jones, J. Liu, and B. Yoshida, Chaos, complexity, and random matrices, JHEP 2017(11), 048

    2017

  3. [3]

    Z.-W. Liu, S. Lloyd, E. Zhu, and H. Zhu, Entangle- ment, quantum randomness, and complexity beyond scrambling, JHEP2018(7)

  4. [4]

    P. Kos, M. Ljubotina, and T. Prosen, Many-body quantum chaos: Analytic connection to random ma- trix theory, Phys. Rev. X8, 021062 (2018)

    2018

  5. [5]

    Huang, R

    H.-Y. Huang, R. Kueng, and J. Preskill, Predicting many properties of a quantum system from very few measurements, Nat. Phys.16, 1050 (2020)

    2020

  6. [6]

    Eisert, D

    J. Eisert, D. Hangleiter, N. Walk, I. Roth, D. Markham, R. Parekh, U. Chabaud, and E. Kashefi, Quantum certification and benchmark- ing, Nat. Rev. Phys.2, 382 (2020)

    2020

  7. [7]

    Elben, S

    A. Elben, S. T. Flammia, H.-Y. Huang, R. Kueng, J. Preskill, B. Vermersch, and P. Zoller, The ran- domized measurement toolbox, Nat. Rev. Phys.5, 9 (2023)

    2023

  8. [8]

    Knill, D

    E. Knill, D. Leibfried, R. Reichle, J. Britton, R. B. Blakestad, J. D. Jost, C. Langer, R. Ozeri, S. Sei- delin, and D. J. Wineland, Randomized benchmark- ing of quantum gates, Phys. Rev. A77, 012307 (2008)

    2008

  9. [9]

    Helsen, I

    J. Helsen, I. Roth, E. Onorati, A. Werner, and J. Eis- ert, General framework for randomized benchmark- ing, PRX Quantum3, 020357 (2022)

    2022

  10. [10]

    Nakata, D

    Y. Nakata, D. Zhao, T. Okuda, E. Bannai, Y. Suzuki, S. Tamiya, K. Heya, Z. Yan, K. Zuo, S. Tamate, Y. Tabuchi, and Y. Nakamura, Quan- tum circuits for exact unitaryt-designs and appli- cations to higher-order randomized benchmarking, PRX Quantum2, 030339 (2021)

    2021

  11. [11]

    Portmann and R

    C. Portmann and R. Renner, Security in quantum cryptography, Rev. Mod. Phys.94, 025008 (2022)

    2022

  12. [12]

    J. S. Cotler, D. K. Mark, H.-Y. Huang, F. Hern´ andez, J. Choi, A. L. Shaw, M. En- dres, and S. Choi, Emergent quantum state designs from individual many-body wave functions, PRX Quantum4, 010311 (2023)

    2023

  13. [13]

    J. Choi, A. L. Shaw, I. S. Madjarov, X. Xie, R. Finkelstein, J. P. Covey, J. S. Cotler, D. K. Mark, H.-Y. Huang, A. Kale,et al., Preparing random states and benchmarking with many-body quantum chaos, Nature613, 468 (2023)

    2023

  14. [14]

    Zhang, P

    B. Zhang, P. Xu, X. Chen, and Q. Zhuang, Holo- graphic deep thermalization for secure and efficient quantum random state generation, Nat. Commun. 16, 6341 (2025)

    2025

  15. [15]

    Pilatowsky-Cameo, I

    S. Pilatowsky-Cameo, I. Marvian, S. Choi, and W. W. Ho, Hilbert-space ergodicity in driven quan- tum systems: Obstructions and designs, Phys. Rev. X14, 041059 (2024)

    2024

  16. [16]

    Gross, K

    D. Gross, K. Audenaert, and J. Eisert, Evenly dis- tributed unitaries: On the structure of unitary de- signs, J. Math. Phys48, 052104 (2007)

    2007

  17. [17]

    Dankert, R

    C. Dankert, R. Cleve, J. Emerson, and E. Livine, Exact and approximate unitary 2-designs and their application to fidelity estimation, Phys. Rev. A80, 012304 (2009)

    2009

  18. [18]

    Emerson, Y

    J. Emerson, Y. S. Weinstein, M. Saraceno, S. Lloyd, and D. G. Cory, Pseudo-random unitary operators for quantum information processing, Science302, 2098 (2003)

    2098

  19. [19]

    F. G. S. L. Brand˜ ao, A. W. Harrow, and M. Horodecki, Local random quantum circuits are approximate polynomial-designs, Commun. Math. Phys346, 397 (2016)

    2016

  20. [20]

    F. G. S. L. Brand˜ ao, A. W. Harrow, and M. Horodecki, Efficient quantum pseudorandomness, Phys. Rev. Lett.116, 170502 (2016)

    2016

  21. [21]

    Haferkamp, Random quantum circuits are approx- imate unitaryt-designs in depthO(nt 5+o(1)), Quan- 7 tum6, 795 (2022)

    J. Haferkamp, Random quantum circuits are approx- imate unitaryt-designs in depthO(nt 5+o(1)), Quan- 7 tum6, 795 (2022)

    2022

  22. [22]

    Schuster, J

    T. Schuster, J. Haferkamp, and H.-Y. Huang, Ran- dom unitaries in extremely low depth, Science389, adv8590 (2025)

    2025

  23. [23]

    Pilatowsky-Cameo, C

    S. Pilatowsky-Cameo, C. B. Dag, W. W. Ho, and S. Choi, Complete Hilbert-space ergodicity in quan- tum dynamics of generalized Fibonacci drives, Phys. Rev. Lett.131, 250401 (2023)

    2023

  24. [24]

    L. Cui, T. Schuster, L. Mao, H.-Y. Huang, and F. Brandao, Random unitaries from hamiltonian dy- namics (2025), arXiv:2510.08434

    arXiv 2025

  25. [25]

    L. Mao, L. Cui, T. Schuster, and H.-Y. Huang, Random unitaries that conserve energy (2025), arXiv:2510.08448

    arXiv 2025

  26. [26]

    Y.-N. Zhou, R. L¨ owenberg, and J. Sonner, Realizing unitaryk-designs with a single quench, Phys. Rev. Lett.136, 220403 (2026)

    2026

  27. [27]

    Zhou, T.-G

    Y.-N. Zhou, T.-G. Zhou, and J. Sonner, Three hamil- tonians are sufficient for unitaryk-design in tempo- ral ensemble (2026), arXiv:2604.04205

    Pith/arXiv arXiv 2026

  28. [28]

    Sun and P

    N. Sun and P. Zhang, Unitary designs from two chaotic Hamiltonians and a random Pauli operation (2026), arXiv:2604.10122

    Pith/arXiv arXiv 2026

  29. [29]

    D. A. Roberts and B. Yoshida, Chaos and complexity by design, JHEP2017(4)

  30. [30]

    D. K. Mark, F. Surace, A. Elben, A. L. Shaw, J. Choi, G. Refael, M. Endres, and S. Choi, Max- imum entropy principle in deep thermalization and in Hilbert-space ergodicity, Phys. Rev. X14, 041051 (2024)

    2024

  31. [31]

    Dowling and K

    N. Dowling and K. Modi, Operational met- ric for quantum chaos and the corresponding spatiotemporal-entanglement structure, PRX Quan- tum5, 010314 (2024)

    2024

  32. [32]

    O’Donovan, N

    P. O’Donovan, N. Dowling, K. Modi, and M. T. Mitchison, Diagnosing chaos with projected ensem- bles of process tensors, PRX Quantum7, 020322 (2026)

    2026

  33. [33]

    Roy and A

    A. Roy and A. J. Scott, Unitary designs and codes, Des. Codes Cryptogr.53, 13 (2009)

    2009

  34. [34]

    von Neumann, Proof of the ergodic theorem and the H-theorem in quantum mechanics, Eur

    J. von Neumann, Proof of the ergodic theorem and the H-theorem in quantum mechanics, Eur. Phys. J. H35, 201 (2010)

    2010

  35. [35]

    Zhang, H

    D. Zhang, H. T. Quan, and B. Wu, Ergodicity and mixing in quantum dynamics, Phys. Rev. E94, 022150 (2016)

    2016

  36. [36]

    Webb, The Clifford group forms a unitary 3- design, Quantum Inf

    Z. Webb, The Clifford group forms a unitary 3- design, Quantum Inf. Comput.16, 1379 (2016)

    2016

  37. [37]

    H. Zhu, R. Kueng, M. Grassl, and D. Gross, The Clif- ford group fails gracefully to be a unitary 4-design, arXiv:1609.08172 (2016)

    Pith/arXiv arXiv 2016

  38. [38]

    Zhu, Multiqubit Clifford groups are unitary 3- designs, Phys

    H. Zhu, Multiqubit Clifford groups are unitary 3- designs, Phys. Rev. A96, 062336 (2017)

    2017

  39. [39]

    Gross, S

    D. Gross, S. Nezami, and M. Walter, Schur–Weyl duality for the Clifford group with applications: Property testing, a robust Hudson theorem, and de Finetti representations, Commun. Math. Phys385, 1325 (2021)

    2021

  40. [40]

    Z. Wang, Y. Wang, and B. Wu, Quantum chaos and physical distance between quantum states, Phys. Rev. E103, 042209 (2021)

    2021

  41. [41]

    Frigg, J

    R. Frigg, J. Berkovitz, and F. Kronz, The ergodic hierarchy, inThe Stanford Encyclopedia of Philos- ophy, edited by E. N. Zalta (Metaphysics Research Lab, Stanford University, Stanford, CA, USA, 2025)

    2025

  42. [42]

    Helsen and M

    J. Helsen and M. Walter, Thrifty shadow estimation: Reusing quantum circuits and bounding tails, Phys. Rev. Lett.131, 240602 (2023)

    2023

  43. [43]

    Chen and H

    D. Chen and H. Zhu, Nonstabilizerness enhances thrifty shadow estimation (2024), arXiv:2410.23977

    Pith/arXiv arXiv 2024

  44. [44]

    Z. Yang, D. Chen, Z. Li, and H. Zhu, High-precision fidelity estimation with common randomized mea- surements (2025), arXiv:2511.22509

    arXiv 2025

  45. [45]

    Elben, B

    A. Elben, B. Vermersch, M. Dalmonte, J. I. Cirac, and P. Zoller, R´ enyi entropies from random quenches in atomic Hubbard and spin models, Phys. Rev. Lett. 120, 050406 (2018)

    2018

  46. [46]

    Vermersch, A

    B. Vermersch, A. Elben, M. Dalmonte, J. I. Cirac, and P. Zoller, Unitaryn-designs via random quenches in atomic Hubbard and spin models: Ap- plication to the measurement of R´ enyi entropies, Phys. Rev. A97, 023604 (2018)

    2018

  47. [47]

    Nandy, Unitaryk-designs without hamiltonian quenches (2026), arXiv:2607.20640

    P. Nandy, Unitaryk-designs without hamiltonian quenches (2026), arXiv:2607.20640

    Pith/arXiv arXiv 2026

  48. [48]

    D. I. Pikulin and M. Franz, Black hole on a chip: Proposal for a physical realization of the Sachdev- Ye-Kitaev model in a solid-state system, Phys. Rev. X7, 031006 (2017)

    2017

  49. [49]

    Asaduzzaman, R

    M. Asaduzzaman, R. G. Jha, and B. Sambasivam, Sachdev-Ye-Kitaev model on a noisy quantum com- puter, Phys. Rev. D109, 105002 (2024)

    2024

  50. [50]

    Cao, Y.-N

    Y. Cao, Y.-N. Zhou, T.-T. Shi, and W. Zhang, Towards quantum simulation of Sachdev-Ye-Kitaev model, Science Bulletin65, 1170 (2020)

    2020

  51. [51]

    Labuhn, D

    H. Labuhn, D. Barredo, S. Ravets, S. de L´ es´ eleuc, T. Macr` ı, T. Lahaye, and A. Browaeys, Tunable two- dimensional arrays of single Rydberg atoms for real- izing quantum Ising models, Nature (London)534, 667 (2016)

    2016

  52. [52]

    Marcuzzi, J

    M. Marcuzzi, J. Min´ aˇ r, D. Barredo, S. de L´ es´ eleuc, H. Labuhn, T. Lahaye, A. Browaeys, E. Levi, and I. Lesanovsky, Facilitation dynamics and localization phenomena in Rydberg lattice gases with position disorder, Phys. Rev. Lett.118, 063606 (2017)

    2017

  53. [53]

    A. G. de Oliveira, E. Diamond-Hitchcock, D. M. Walker, M. T. Wells-Pestell, G. Pelegr´ ı, C. J. Picken, G. P. A. Malcolm, A. J. Daley, J. Bass, and J. D. Pritchard, Demonstration of weighted-graph opti- mization on a Rydberg-atom array using local light shifts, PRX Quantum6, 010301 (2025). 8 Appendix A: Typicality of Theorem 1 Although Theorem 1 is stated...

    2025

  54. [54]

    Haar twirling and Leading-Order Analysis 2 S3

    Proof of Theorem S1 2 S2. Haar twirling and Leading-Order Analysis 2 S3. Proof of Theorem 1 3

  55. [55]

    Proof of Corollary S1 5

  56. [56]

    Proof of Proposition S1 5

  57. [57]

    Proof of Proposition S2 6

  58. [58]

    Proof of Proposition S3 6

  59. [59]

    Proof of Proposition S4 7

  60. [60]

    Frame potential ofE L(H) 10 In this Supplemental Material, we provide the technical details supporting the results presented in the main text

    Proof of Theorem 1 8 S4. Frame potential ofE L(H) 10 In this Supplemental Material, we provide the technical details supporting the results presented in the main text. We first give a more formal discussion and proof of the time-averaging procedure under thek-th order additive nonresonance condition. We then review the Haar twirling channel and its leadin...

  61. [61]

    Proof of Theorem S1 Proof of Theorem S1.ForU, Vdefined in Eq. (S1), we have ⏐⏐Tr(V†U) ⏐⏐2 = ⏐⏐Tr ( Q† e−iHδt 2Pe−iHδt 1 )⏐⏐2 = ∑ m1,n1 ∑ ˜m1,˜n1 ⟨n1|Q†|m1⟩⟨m1|P|n 1⟩⟨˜n1|P†|˜m1⟩⟨˜m1|Q|˜n1⟩e−iδt1(En1−E ˜n1)−iδt2(Em1−E ˜m1) = ∑ m1,n1 ∑ ˜m1,˜n1 Pm1n1Q† n1m1P† ˜n1 ˜m1Q ˜m1˜n1 e−iδt1(En1−E ˜n1)−iδt2(Em1−E ˜m1).(S7) Herem 1, n1,˜m1,˜n1 are eigenstate labels ofH...

  62. [62]

    (S16) and (S13) into Eq

    Proof of Corollary S1 Proof of Corollary S1.The average of R π,σ(P,Q) over a Haar-random basis is EW Rπ,σ(P,Q) =E W ∑ m,n∈[Dk] 1 S(m)S(n) Tr [ ( P⊗Q †⊗P†⊗Q )⊗k k∏ r=1 ( W†|mr⟩⟨nr|W⊗W †|nr⟩⟨mr|W ⊗W†|nσ(r)⟩⟨mπ(r)|W⊗W †|mπ(r)⟩⟨nσ(r)|W )] = ∑ m,n∈[Dk] 1 S(m)S(n) Tr [ ( P⊗Q †⊗P†⊗Q )4k Φ4k [k∏ r=1 ( |mr⟩⟨nr|⊗|n r⟩⟨mr| ⊗|nσ(r)⟩⟨mπ(r)|⊗|m π(r)⟩⟨nσ(r)| )] .(S29) S...

  63. [63]

    This proves Proposition S1

    Proof of Proposition S1 Proof of Proposition S1.Since the ensembleEsatisfies F (k) E =E U,V∼E ⏐⏐Tr(V†U) ⏐⏐2k =O k(D2k−ϵk), ϵ k = Ωk(1).(S35) By H¨ older’s inequality, for 1≤ℓ≤k, we have F (ℓ) E =E U,V∼E ⏐⏐Tr(V†U) ⏐⏐2ℓ ≤ ( EU,V∼E ⏐⏐Tr(V†U) ⏐⏐2k)ℓ/k ≤O ℓ(D2ℓ−ϵℓ),(S36) whereϵ ℓ =ℓϵ k/k= Ω ℓ(1) is still lower bounded by a constant with respect to dimensionDwi...

  64. [64]

    Proof of Proposition S2 Proof of Proposition S2.Sinceµ i defined in Eq. (S4) counts the number of occurrences of labeliinm, for any nonnegative integer vector (µ 1,...,µ D) satisfyingµ 1 +···+µ D =k, the number of vectorsm∈[D] k with these occupation numbers is k! µ1!µ2!···µ D!.(S37) Therefore the total contribution of all vectors with the same occupation...

  65. [65]

    Its element is a product of Kronecker delta functions

    Proof of Proposition S3 Proof of Proposition S3.Since Gπ,σ(β) = ∑ m,n∈[D]k ⟨x(m,n)|Vβ|y(m,n)⟩ S(m)S(n) .(S41) The matrix element ⟨x(m,n)|Vβ|y(m,n)⟩(S42) is nonzero exactly when the indices ofx(m,n) match those ofV βy(m,n). Its element is a product of Kronecker delta functions. Hence its nonzero condition imposes only equality constraints among the variabl...

  66. [66]

    Lets=s p +sq be the total number of fixed points ofα, wheres p ands q denote the numbers of fixed points associated withP-type andQ-type factors, respectively

    Proof of Proposition S4 Proof of Proposition S4.We first prove that ifαhas a fixed point, equivalently a 1-cycle, then its contribution is suppressed. Lets=s p +sq be the total number of fixed points ofα, wheres p ands q denote the numbers of fixed points associated withP-type andQ-type factors, respectively. Using the cycle expansion of permutation trace...

  67. [67]

    For a permutationη∈S k, we write y|Fix(η)| = k∏ i=1 [ 1 + (y−1)δ η(i)=i ] .(S52) 8 We now expand this product

    Proof of Lemma S1 Proof of Lemma S1.Let [k] ={1,...,k}. For a permutationη∈S k, we write y|Fix(η)| = k∏ i=1 [ 1 + (y−1)δ η(i)=i ] .(S52) 8 We now expand this product. In the expansion, for each positioni∈[k] one chooses either the term 1 or the term (y−1)δ η(i)=i. Equivalently, one chooses a subsetA⊆[k] of positions from which the second term is selected....

  68. [68]

    Proof of Theorem 1.First, we define the subsetM 4k of the permutation groupS 4k to be the set that contains all permutations that consist of 2kdisjoint transpositions

    Proof of Theorem 1 Prepared with the results above, we are ready to prove Theorem 1. Proof of Theorem 1.First, we define the subsetM 4k of the permutation groupS 4k to be the set that contains all permutations that consist of 2kdisjoint transpositions. By Eq. (S14) and Proposition S3, we have Rπ,σ = ∑ α,β∈S4k Wg(4k) D (α−1β)EP,Q∼Eint[Cα(P,Q)] Gπ,σ(β) = ∑ ...