Pith. sign in

REVIEW 2 major objections 4 minor 61 references

A single filtering commutator can reshape a quantum control algebra to match any chosen subalgebra, while projectors compose distinct systems and two indices diagnose invariance.

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

2026-08-02 18:44 UTC pith:D2ROUUJQ

load-bearing objection A clean composition theorem and a plausible nearest-neighbor generating set, but the paper's central reduction theorem is false as stated — the counterexample kills the unified-framework claim. the 2 major comments →

arxiv 2603.04916 v2 pith:D2ROUUJQ submitted 2026-03-05 quant-ph

A Dynamical Lie-Algebraic Framework for Hamiltonian Engineering and Quantum Control

Yanying Liang , Ruibin Xu , Mao-Sheng Li , Haozhen Situ , Zhu-Jun Zheng This is my paper
classification quant-ph
keywords dynamical Lie algebraHamiltonian engineeringquantum controlgenerator-set reductionfiltering operatordirect-sum compositionPauli stringsIsing model simulation
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.

The paper tries to turn dynamical Lie algebras (DLAs) from a descriptive tool into a design tool for quantum control and Hamiltonian engineering. It answers three questions about editing a set of available Hamiltonians: how to combine different systems into one qubit-efficient direct sum, when adding or reorganizing terms leaves the reachable algebra unchanged, and how to reduce a large algebra to a chosen subalgebra. The central reduction claim is that taking commutators with a single filtering operator that lives in the target subalgebra and is non-central in each of its simple parts produces a new generator set whose Lie closure is exactly that target subalgebra. If correct, this gives a principled way to engineer evolutions that stay inside a symmetry sector or a low-dimensional subspace, which matters for circuit pruning, simulation, and trainability of variational circuits.

Core claim

Theorem 3 is the anchor: for a cyclic reductive DLA g_A = ⊕_{j=1}^T g_j, with target subalgebra h = ⊕_{j=1}^S g_j, choose F ∈ h whose projection onto each g_j (j ≤ S) is non-zero and non-central. Then the filtered generating set A' = {[F, A] : A ∈ A} satisfies g_{A'} ≅ h. The mechanism is that the unwanted components g_{S+1},...,g_T commute with F and are killed, while within each retained simple ideal the non-central projection F_j acts as a seed whose commutators generate all of g_j. The paper also proves a direct-sum composition theorem using mutually orthogonal spectral projectors on an auxiliary register, and gives two quantitative diagnostics — projection overlap and DLA percentage cha

What carries the argument

The dynamical Lie algebra (DLA), the closure under nested commutators of the available anti-Hermitian generators, is the object being edited. The reduction machinery is the filtering operator F: an element of the target subalgebra whose non-central projections seed each retained simple ideal; the classically computed commutators [F, A] filter the generator set. Composition uses mutually orthogonal spectral projectors Π_m of an auxiliary Hermitian operator to keep distinct DLAs independent on a shared register. Invariance uses the projection-overlap index O(P,Q), measuring how much a candidate operator's central projection lies within the existing central span, together with the relative dime

Load-bearing premise

The reduction theorem's load-bearing step is the assumption that a non-central filter operator inside any one building block of the algebra generates that entire building block even when only the original Hamiltonians' projections are used as partners; the proof validates this for the full block, not for the restricted projected set, and if that step fails the filtered generators may span only part of the target.

What would settle it

Take a cyclic generating set for g_A = su(2) ⊕ su(2) on two qubits, set h to the first su(2), choose F = iZ_1, form A' = {[F,A]}, and compute dim⟨A'⟩. If any such example gives dimension less than 3, Theorem 3's projection-step assumption fails; the paper's harmonic-oscillator example does not test this step because the projected generator set there already contains both X_j and P_j in each mode.

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

Share X Bluesky LinkedIn Reddit HN

If this is right

  • If Theorem 3 holds, any cyclic reductive DLA can be reduced to any ideal-sum subalgebra by classically computing single commutators with F, with no ancilla and no physical simulation of F required.
  • DLA composition via spectral projectors lets K distinct subsystem DLAs run in parallel on the same qubits plus ⌈log K⌉ ancillas, rather than K separate copies of the system.
  • The nearest-neighbor generating construction for su(2^N) gives a hardware-friendly 2N+1 generator set that preserves full controllability after reorganization.
  • The projection-overlap and percentage-change indices give a quantitative pruning criterion: generators with O≈1 and D_c≈0 are algebraically redundant and can be removed without changing the reachable algebra.
  • The LTFIM-to-TFIM error bound, ∥U_LTFIM(t) − U_TFIM(t)∥ ≤ C n^2 α^2 t^2, shows that approximate subalgebra reduction can be certified even in a non-cyclic example.
  • The reduction construction gives a systematic route to engineering evolutions confined to decoherence-free subspaces and low-dimensional symmetry sectors, addressing a core obstacle in near-term quantum simulation.

Where Pith is reading between the lines

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

  • The filtering construction suggests a general recipe for Hamiltonian pruning by symmetry: any single non-central element inside a target ideal sum can serve as F, which may extend naturally to systems beyond the paper's examples, such as fermionic or lattice models with known ideal decompositions.
  • The two indices O and D_c could be used as early trainability diagnostics in variational circuits: because DLA dimension is tied to barren plateaus, monitoring these quantities during ansatz growth may give a warning before full Lie closure is computed.
  • Theorem 1 could be iterated to build hierarchical tensor-product direct sums, leading to recursive parallel-simulation schemes where each block is itself a composed DLA.
  • A direct testable extension is to apply the F-filtering reduction to random cyclic Pauli-string generating sets for su(2)⊕su(2) and check whether g_{A'} exactly equals the target ideal; failures would localize exactly where the proof's projection-step assumption breaks.

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 / 4 minor

Summary. The manuscript develops a dynamical-Lie-algebra (DLA) framework for three generator-set operations: composition, invariance, and reduction. Theorem 1 constructs qubit-efficient direct sums of distinct DLAs via spectral projectors on an ancillary register. Theorem 2 and the surrounding section give Pauli-string constructions and two quantitative indices (projection overlap O and percentage change D_c) for approximate DLA invariance. Theorem 3 claims that, for a cyclic reductive DLA g_A=⊕_j g_j and a target subalgebra h=⊕_{j≤S}g_j, the filtered generating set A'={[F,A]: A∈A}, with F∈h having non-central projection on each g_j, satisfies g_{A'}≅h. The paper also presents numerical examples, including a harmonic-oscillator illustration and an LTFIM-to-TFIM error bound. The main novelty is the reduction result, presented as the answer to the paper's Question 3.

Significance. If valid, Theorem 1 and Theorem 2 would be useful additions: the direct-sum construction is explicit and appears sound, and the su(2^N) nearest-neighbour generator construction is a plausible improvement for hardware-aware circuit compilation. The paper is also commendable for stating clear algebraic questions and providing worked examples. However, the central reduction theorem, Theorem 3, is false as stated: there is a simple su(2) counterexample satisfying all hypotheses. Since Theorem 3 is the paper's main answer to Question 3 and anchors the claimed reduction framework, this falsification is load-bearing. The LTFIM error-bound derivation in Appendix C2 also contains an inconsistency. The paper should not be accepted in its current form.

major comments (2)
  1. [§III.C, Theorem 3 and Appendix C1, Eq. (C3)] The surjectivity step of Theorem 3 is invalid, and the statement is false. Consider g_A=su(2) with the cyclic generating set A={iσ_x, iσ_z}. It is cyclic: for C=[iσ_z,iσ_x]=2iσ_y, one has [iσ_x,[iσ_x,C]]∝C. Let h=su(2) and F=iσ_z, which is non-central. Then A'={[F,A]}={2iσ_y,0}, so g_{A'}=span_R{iσ_y}≅u(1), not su(2). The proof after Eq. (C3) invokes Proposition 2 for the restricted set {[F_j,π_j(A)]:A∈A}, but Proposition 2 only applies to {[F_j,Y]:Y∈g_j}. The 'therefore' step, from the existence of some Y∈g_j with [F_j,Y]≠0 to the conclusion that the projected generators generate g_j, is not justified. The su(2) example shows the claim fails even when π_j(A)=A generates g_j. Adding a commuting su(2) factor and taking F in one factor gives a proper-subalgebra version. Thus Theorem 3 (and Theorem A3) is false.
  2. [Appendix C2, Eqs. (C4)-(C8)] The derivation of the LTFIM-to-TFIM error bound is internally inconsistent. The claim that H̃_γ^1(τ)=e^{iτH0}σ^z_γ e^{-iτH0} remains a single-qubit operator is actually correct, since σ^z_γ commutes with all σ^z_jσ^z_{j+1} terms and only the local σ^x_γ term contributes to the adjoint action. However, after correctly observing that [H̃_{γ1}(s), H̃_{γ2}(ν)]=0 for γ1≠γ2, the text replaces the double sum by a constant C′≈n² 'accounting for the number of non-zero commutator pairs.' The sum is zero, so Lemma A1's bound is zero; the n² factor is not derived. The final inequality (Eq. (20)) may be a valid loose bound, but the presented derivation does not establish it.
minor comments (4)
  1. [§IV.C, harmonic oscillators] This example lies outside the finite-dimensional hypothesis of Theorem 3: each mode is an infinite-dimensional Fock space. Moreover, the algebra generated by {iX_j,iP_j,iN_j,iS_j} contains a central element iI_j, so calling each g_j a simple Lie subalgebra is not accurate. The example should be reframed as an analogy rather than an application of Theorem 3.
  2. [Appendix C2] The sentence 'For γ1=γ2, the commutator is trivially zero' is not correct as a general statement: two Heisenberg-evolved operators of the same single-qubit σ^z at different times do not necessarily commute. This case is not in the sum over γ1<γ2 and should simply be omitted.
  3. [Eq. (10)] The notation for the product over 2<j<i is garbled ('Y' instead of a tensor product symbol) and the expression is hard to parse. Please rewrite with explicit tensor products or define the string ordering.
  4. [§IV.B, central-spin example] The projection overlap O(P,Q) is introduced for Pauli DLAs, but the example uses a central basis including iI_8. Clarify how the trace normalization and the orthonormal basis are chosen in this non-Pauli central space.

Circularity Check

0 steps flagged

No significant circularity: the paper's derivations are independent of their conclusions; the main risk is a correctness gap in Appendix C, not circularity.

full rationale

The paper's three main constructions are not circular. Theorem 1 is an independent extension of Lemma 1: it verifies [A⊗Π_i, B⊗Π_j]=0 for i≠j and constructs an explicit isomorphism φ_m(X)=X⊗Π_m; the conclusion g_{A'}≅⊕_m g_{A_m} is not assumed in the definition of A'. Theorem 2 relies on Proposition 1 (direct commutator computation) and external Lemma 2 (Smith et al.), and the Appendix B induction produces B'_I from B'_II by explicit nested brackets. The invariance indices in Sec. III.B.2 are definitions rather than predictions: O(P,Q)=1 is explicitly acknowledged to be a sufficient condition for Lemma 3's Condition 2, and D_c is the tautological relative dimension change, so no fitted parameter is later relabeled as a prediction. Theorem 3's reduction is the only place with a load-bearing inference: the step marked 'Therefore' in Appendix C asserts that the restricted set A'_j={[F_j,A_h^{(j)}]: A∈A} generates g_j because the full commutator set {[F_j,Y]: Y∈g_j} does (Proposition 2). This is a logical gap, and the paper's own cyclic-limitation paragraph acknowledges that non-cyclic cases fail; the cyclic hypothesis only ensures the restricted set is nonempty, not that it Lie-generates the ideal. But this is a correctness/falsifiability problem, not circularity: Proposition 2 is an independent lemma with its own proof, and the theorem's conclusion is not used as a premise. The error bound for LTFIM/TFIM uses the external product-formula bound of Bosse et al. and simplifies commutators; any error in estimating C' is an algebraic mistake, not a circular reduction. There are no self-citations by the current authors in the load-bearing chain. Therefore no Eq. X is equivalent to Eq. Y by construction, and the score is 0.

Axiom & Free-Parameter Ledger

0 free parameters · 9 axioms · 0 invented entities

The composition and invariance results rest on standard Lie algebra and external lemmas; the reduction and error-bound results depend on assumptions that are either unproven or false. No numbers are fit to data.

axioms (9)
  • standard math Reductive DLA admits direct-sum decomposition into simple ideals (Eq 15)
    Used in Theorem 3; standard structure theorem for reductive Lie algebras [36].
  • standard math For a simple Lie algebra, the roots non-orthogonal to a nonzero Cartan element generate the whole root system (Prop 2 proof)
    Needed to conclude ⟨[F,X]⟩_{Lie}=g; the proof invokes irreducibility/connected Dynkin diagram but does not give a full argument for all real forms.
  • standard math Cartan subalgebras of a simple complex Lie algebra are conjugate under inner automorphisms
    Used in Prop 2 to reduce semisimple F to a fixed Cartan subalgebra.
  • domain assumption Skew-Hermitian generators make every filtering operator F diagonalizable/semisimple
    Appendix C states this is naturally satisfied for quantum control DLAs; it is true for skew-Hermitian matrices but the paper does not justify the real-vs-complex root-space setup.
  • domain assumption Theorem 3 requires the generating set A to be cyclic (Eq 6)
    The theorem is stated only for cyclic reductive DLAs; cyclicity is not actually used in the surjectivity step beyond asserting a non-zero commutator exists, which is part of the gap.
  • domain assumption Lemma 2 of Smith et al. [23]: minimal Pauli su(2^N) generating sets and product-universal subsets
    Black-box external result used in Theorem 2 and its proof.
  • domain assumption Lemma 3 of Zimborás et al. [22]: necessary and sufficient simulability conditions
    Used as benchmark for the O(P,Q)/D_c indices; the paper criticizes its binary nature but uses its condition.
  • domain assumption Lemma A1 of Bosse et al. [31]: product-formula error bound for H=H0+αH1
    Used to derive Eq (20); the application in Appendix C2 misstates the Heisenberg-evolved operators as single-qubit and disjoint.
  • ad hoc to paper Heisenberg evolution of σ^z_γ under TFIM remains a single-qubit operator (and distinct γ operators act on disjoint qubits)
    This is asserted in Appendix C2 and is false for nearest-neighbor H0; support spreads in time, so commutators between distinct γ generically do not vanish.

pith-pipeline@v1.3.0-alltime-deepseek · 21196 in / 22028 out tokens · 192494 ms · 2026-08-02T18:44:54.630142+00:00 · methodology

0 comments
Cite this review

Pith. "Pith review of A Dynamical Lie-Algebraic Framework for Hamiltonian Engineering and Quantum Control." pith.science (2026). https://pith.science/paper/D2ROUUJQ

@misc{pith2026260304916,
  author       = {Pith},
  title        = {Pith review of: A Dynamical Lie-Algebraic Framework for Hamiltonian Engineering and Quantum Control},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/D2ROUUJQ}},
  note         = {Machine review of arXiv:2603.04916}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Determining the unitary dynamics accessible from finite Hamiltonian resources is a central problem in Hamiltonian engineering and quantum control. Dynamical Lie algebras (DLAs) connect available control Hamiltonians with the reachable dynamics, but their use as a design tool for modifying Hamiltonian generator sets remains less developed. In this work, we develop a finite-dimensional DLA framework for three generator-set operations: composition, invariance, and reduction. For composition, we construct direct sums of component DLAs using spectral projectors on an auxiliary register. For invariance, we analyze when modifications of Pauli-string generating sets preserve the generated Lie algebra, and introduce algebraic diagnostics for added generators. For reduction, we consider compact reductive DLAs and show how projection onto selected simple ideals gives reduced generating sets whose Lie closures are the corresponding ideal sums. We illustrate these results with finite-dimensional examples and numerical checks, including direct-sum dimension addition, central-spin invariance diagnostics, and DLA-based ansatz reduction for block-local Hamiltonians. The results show how DLA structure can be used to diagnose controllability and guide Hamiltonian generator design under explicit algebraic assumptions.

Figures

Figures reproduced from arXiv: 2603.04916 by Haozhen Situ, Mao-Sheng Li, Ruibin Xu, Yanying Liang, Zhu-Jun Zheng.

Figure 1
Figure 1. Figure 1: FIG. 1: Illustration of three questions on modifying the generating set [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2: Lie-theoretic background. The evolution of a quantum state [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3: Illustration of Theorem 1. (a) shows the composition of generator sets for the dipole–field coupling and [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4: Comparison of dynamical evolution versus algebraic structural stability under different perturbations. [PITH_FULL_IMAGE:figures/full_fig_p008_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5: Schematic illustration of the generator sets for [PITH_FULL_IMAGE:figures/full_fig_p010_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6: Guiding physical circuit design via theoretical DLA modifications. [PITH_FULL_IMAGE:figures/full_fig_p010_6.png] view at source ↗

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

61 extracted references · 5 linked inside Pith

  1. [1]

    Lemma 2 tells us that we can modify the generating setA → A′ while preservingsu(2 N )with no increasing of the cardinalities

    Generating set of Pauli strings with the cardinality unchanged WhenL=L ′, a special case we choose is the full DLA su(2N ). Lemma 2 tells us that we can modify the generating setA → A′ while preservingsu(2 N )with no increasing of the cardinalities. LetAbe a minimal universal generating set forsu(2 N )withL= 2N+ 1, andA ⊂iP ∗ N . Here,P ∗ N denotes all Pa...

  2. [2]

    Generating set of Pauli strings with increasing of the cardinality Unlike the strict invariance in theL=L ′ case, forL < L′, we propose two quantitative metrics to evaluate approximate invariance, thereby overcoming the limitations in Condition (2) of Lemma 3. Lemma 3 provides two necessary and sufficient conditions for the above question preserving⟨P⟩ Li...

  3. [3]

    For each operator inP, usingT r(ρ † α ·iH β), we computeΠ(iH 1) = [0,1] T andΠ(iH 2) = [0,2] T , which are two linearly dependent vectors

    Here J2 ∈Rbeing the interaction strengths. For each operator inP, usingT r(ρ † α ·iH β), we computeΠ(iH 1) = [0,1] T andΠ(iH 2) = [0,2] T , which are two linearly dependent vectors. Thus, the central projection span space ofPis span(Π(P)) =span{ρ 2}, and the rank of the projection ma- trix ˜Tisrank( ˜T) = 1. AddQ 1 ={iH 3}whereiH 3 =i(2σ z 1 + 0.5σx 1 σx ...

  4. [4]

    DLA Reduction ,21 ,,, LieLiHiHiH g iHtiHt eet 0)(   Hamiltonian Learning       ？ ？ l l l L l l HHH  1 1.Parallel Architecture

  5. [5]

    Compiler Optimization

  6. [6]

    6: Guiding physical circuit design via theoretical DLA modifications

    Circuit Pruning iHtetU )(U(t) Hamiltonian Simulation for a given H Unitary Dynamics Find the true Hamiltonian generating unitary dynamics H FIG. 6: Guiding physical circuit design via theoretical DLA modifications. the ratio of the weak longitudinal field to the dominant inter- action strength. The details to derive this error bound is in Appendix C 2....

  7. [7]

    Fring and M

    A. Fring and M. H. Moussa, Unitary quantum evolution for time-dependent quasi-Hermitian systems with nonobservable Hamiltonians, Phys. Rev. A93, 042114 (2016)

    2016

  8. [8]

    Y . Lai, J. Liang, H. Müller-Kirsten, and J.-G. Zhou, Time evolu- tion of quantum systems with time-dependent Hamiltonian and the invariant Hermitian operator, J. Phys. A: Math. Gen.29, 1773 (1996)

    1996

  9. [9]

    Y . Cao, S. Jin, and N. Liu, Quantum simulation for time- dependent Hamiltonians-with applications to non-autonomous ordinary and partial differential equations, J. Phys. A: Math. Theor.58, 155304 (2025)

    2025

  10. [10]

    Y . Cao, S. Jin, and N. Liu, Unifying framework for quantum simulation algorithms for time-dependent Hamiltonian dynam- ics, Phys. Rev. Res.7, 043186 (2025)

    2025

  11. [11]

    Pal and K

    K. Pal and K. Pal, Time-dependent Hamiltonians and geome- try of operators generated by them, Phys. Rev. E111, 014104 (2025)

    2025

  12. [12]

    Mizuta and K

    K. Mizuta and K. Fujii, Optimal Hamiltonian simulation for time-periodic systems, Quantum7, 962 (2023)

    2023

  13. [13]

    Ramakrishna and H

    V . Ramakrishna and H. Rabitz, Relation between quantum com- puting and quantum controllability, Phys. Rev. A54, 1715 (1996)

    1996

  14. [14]

    Gago-Encinas, T

    F. Gago-Encinas, T. Hartung, D. M. Reich, K. Jansen, and C. P. Koch, Determining the ability for universal quantum comput- ing: Testing controllability via dimensional expressivity, Quan- tum7, 1214 (2023)

    2023

  15. [15]

    D’Alessandro,Introduction to quantum control and dynam- ics(Chapman and Hall/CRC, 2021)

    D. D’Alessandro,Introduction to quantum control and dynam- ics(Chapman and Hall/CRC, 2021)

    2021

  16. [16]

    D’Alessandro and Y

    D. D’Alessandro and Y . Isik, Controllability of the periodic quantum Ising spin chain and the Onsager algebra, J. Phys. A: Math. Theor.58, 115202 (2025)

    2025

  17. [17]

    Schirmer, I

    S. Schirmer, I. Pullen, and A. Solomon, Identification of dy- namical Lie algebras for finite-level quantum control systems, J. Phys. A: Math. Gen.35, 2327 (2002)

    2002

  18. [18]

    Ragone, B

    M. Ragone, B. N. Bakalov, F. Sauvage, A. F. Kemper, C. Or- tiz Marrero, M. Larocca, and M. Cerezo, A Lie algebraic the- ory of barren plateaus for deep parameterized quantum circuits, Nat. Commun.15, 7172 (2024)

    2024

  19. [19]

    Fontana, D

    E. Fontana, D. Herman, S. Chakrabarti, N. Kumar, R. Yalovet- zky, J. Heredge, S. H. Sureshbabu, and M. Pistoia, Characteriz- ing barren plateaus in quantum ansätze with the adjoint repre- sentation, Nat. Commun.15, 7171 (2024)

    2024

  20. [20]

    J. R. McClean, S. Boixo, V . N. Smelyanskiy, R. Babbush, and H. Neven, Barren plateaus in quantum neural network training landscapes, Nat. Commun.9, 4812 (2018)

    2018

  21. [21]

    Larocca, S

    M. Larocca, S. Thanasilp, S. Wang, K. Sharma, J. Bia- monte, P. J. Coles, L. Cincio, J. R. McClean, Z. Holmes, and M. Cerezo, Barren plateaus in variational quantum computing, Nat. Rev. Phys. , 1 (2025)

    2025

  22. [22]

    Wiebe, C

    N. Wiebe, C. Granade, C. Ferrie, and D. G. Cory, Hamiltonian learning and certification using quantum resources, Phys. Rev. Lett.112, 190501 (2014)

    2014

  23. [23]

    Aguilar, S

    G. Aguilar, S. Cichy, J. Eisert, and L. Bittel, Full classification of Pauli Lie algebras, arXiv:2408.00081 (2024)

    Pith/arXiv arXiv 2024

  24. [24]

    Kökcü, R

    E. Kökcü, R. Wiersema, A. F. Kemper, and B. N. Bakalov, Clas- sification of dynamical Lie algebras generated by spin interac- tions on undirected graphs, arXiv:2409.19797 (2024)

    Pith/arXiv arXiv 2024

  25. [25]

    S. Kazi, M. Larocca, M. Farinati, P. J. Coles, M. Cerezo, and R. Zeier, Analyzing the quantum approximate optimization al- gorithm: Ansätze, symmetries, and Lie algebras, PRX Quan- tum6, 040345 (2025)

    2025

  26. [26]

    Rabinovich, A

    D. Rabinovich, A. Kardashin, and S. Adhikary, On the role of overparametrization in quantum approximate optimization, arXiv:2508.10086 (2025)

    Pith/arXiv arXiv 2025

  27. [27]

    Allcock, M

    J. Allcock, M. Santha, P. Yuan, and S. Zhang, On generat- ing direct powers of dynamical Lie algebras, arXiv:2506.05733 (2025)

    Pith/arXiv arXiv 2025

  28. [28]

    Zimborás, R

    Z. Zimborás, R. Zeier, T. Schulte-Herbrüggen, and D. Burgarth, Symmetry criteria for quantum simulability of effective interac- tions, Phys. Rev. A92, 042309 (2015)

    2015

  29. [29]

    I. D. Smith, M. Cautrès, D. T. Stephen, and H. Poulsen Nautrup, Optimally generatingsu(2 n)using Pauli strings, Phys. Rev. Lett.134, 200601 (2025)

    2025

  30. [30]

    Barratt, J

    F. Barratt, J. Dborin, M. Bal, V . Stojevic, F. Pollmann, and A. G. Green, Parallel quantum simulation of large systems on small NISQ computers, npj Quantum Inf.7, 79 (2021)

    2021

  31. [31]

    N. Diaz, P. Braccia, M. Larocca, J. Matera, R. Rossignoli, and M. Cerezo, Parallel-in-time quantum simulation via Page and Wootters quantum time, Phys. Rev. Res.7, 033294 (2025)

    2025

  32. [32]

    Y . Nam, N. J. Ross, Y . Su, A. M. Childs, and D. Maslov, Auto- mated optimization of large quantum circuits with continuous parameters, npj Quantum Inf.4, 23 (2018)

    2018

  33. [33]

    R. Iten, R. Moyard, T. Metger, D. Sutter, and S. Woerner, Exact and practical pattern matching for quantum circuit optimiza- tion, ACM Trans. Quantum Comput.3, 1 (2022)

    2022

  34. [34]

    Viswanathan, O

    Y . Viswanathan, O. Adjoua, C. Feniou, S. Badreddine, and J.-P. Piquemal, An optimal framework for constructing Lie- algebra generator pools: application to variational quantum 12 eigensolvers for chemistry, arXiv:2511.22593 (2025)

    Pith/arXiv arXiv 2025

  35. [35]

    D. A. Lidar, I. L. Chuang, and K. B. Whaley, Decoherence free subspaces for quantum computation, Phys. Rev. Lett.81, 2594 (1998)

    1998

  36. [36]

    G. Xu, J. Zhang, D. Tong, E. Sjöqvist, and L. Kwek, Nona- diabatic holonomic quantum computation in decoherence-free subspaces, Phys. Rev. Lett.109, 170501 (2012)

    2012

  37. [37]

    J. L. Bosse, A. M. Childs, C. Derby, F. M. Gambetta, A. Mon- tanaro, and R. A. Santos, Efficient and practical Hamiltonian simulation from time-dependent product formulas, Nat. Com- mun.16, 2673 (2025)

    2025

  38. [38]

    N. M. Linke, D. Maslov, M. Roetteler, S. Debnath, C. Fig- gatt, K. A. Landsman, K. Wright, and C. Monroe, Experimen- tal comparison of two quantum computing architectures, Proc. Natl. Acad. Sci. U.S.A.114, 3305 (2017)

    2017

  39. [39]

    Zulehner, A

    A. Zulehner, A. Paler, and R. Wille, An efficient methodol- ogy for mapping quantum circuits to the IBM QX architectures, IEEE Trans. Comput.-Aided Des. Integr. Circuits Syst.38, 1226 (2018)

    2018

  40. [40]

    Kulshrestha, X

    A. Kulshrestha, X. Liu, H. Ushijima-Mwesigwa, B. Bach, and I. Safro, QAdaPrune: Adaptive parameter pruning for training variational quantum circuits, in2024 IEEE Int. Conf. Quantum Comput. Eng. (QCE), V ol. 2 (IEEE, 2024) pp. 120–125

    2024

  41. [41]

    Larocca, N

    M. Larocca, N. Ju, D. García-Martín, P. J. Coles, and M. Cerezo, Theory of overparametrization in quantum neural networks, Nat. Comput. Sci.3, 542 (2023)

    2023

  42. [42]

    J. E. Humphreys,Introduction to Lie algebras and representa- tion theory, V ol. 9 (Springer Science & Business Media, 2012)

    2012

  43. [43]

    S. L. Braunstein and P. Van Loock, Quantum information with continuous variables, Rev. Mod. Phys.77, 513 (2005)

    2005

  44. [44]

    Weedbrook, S

    C. Weedbrook, S. Pirandola, R. García-Patrón, N. J. Cerf, T. C. Ralph, J. H. Shapiro, and S. Lloyd, Gaussian quantum informa- tion, Rev. Mod. Phys.84, 621 (2012)

    2012

  45. [45]

    Stinchcombe, Ising model in a transverse field

    R. Stinchcombe, Ising model in a transverse field. I. Basic the- ory, J. Phys. C: Solid State Phys.6, 2459 (1973)

    1973

  46. [46]

    S. I. Mohtashim, A. Das, T. Chatterjee, and F. T. Chowdhury, A near-term quantum simulation of the transverse field Ising model hints at glassy dynamics, Eur. Phys. J. Spec. Top. , 1 (2025)

    2025

  47. [47]

    O. d. A. Bonfim, B. Boechat, and J. Florencio, Ground-state properties of the one-dimensional transverse Ising model in a longitudinal magnetic field, Phys. Rev. E99, 012122 (2019)

    2019

  48. [48]

    M. L. Goh, M. Larocca, L. Cincio, M. Cerezo, and F. Sauvage, Lie-algebraic classical simulations for quantum computing, Phys. Rev. Res.7, 033266 (2025)

    2025

  49. [49]

    Tindall and D

    J. Tindall and D. Sels, Confinement in the transverse field Ising model on the heavy hex lattice, Phys. Rev. Lett.133, 180402 (2024)

    2024

  50. [50]

    R. D. Somma, R. King, R. Kothari, T. E. O’Brien, and R. Bab- bush, Shadow hamiltonian simulation, Nature Communications 16, 2690 (2025)

    2025

  51. [51]

    Q. Zhao, Y . Zhou, A. F. Shaw, T. Li, and A. M. Childs, Hamil- tonian simulation with random inputs, Physical Review Letters 129, 270502 (2022)

    2022

  52. [52]

    Q. Zhao, Y . Zhou, and A. M. Childs, Entanglement accelerates quantum simulation, Nature Physics21, 1338 (2025)

    2025

  53. [53]

    Chakraborty, Implementing any linear combination of uni- taries on intermediate-term quantum computers, Quantum8, 1496 (2024)

    S. Chakraborty, Implementing any linear combination of uni- taries on intermediate-term quantum computers, Quantum8, 1496 (2024)

    2024

  54. [54]

    J. Wang, S. Paesani, R. Santagati, S. Knauer, A. A. Gentile, N. Wiebe, M. Petruzzella, J. L. O’brien, J. G. Rarity, A. Laing, et al., Experimental quantum Hamiltonian learning, Nat. Phys. 13, 551 (2017)

    2017

  55. [55]

    A. Gu, L. Cincio, and P. J. Coles, Practical Hamiltonian learning with unitary dynamics and Gibbs states, Nat. Commun.15, 312 (2024)

    2024

  56. [56]

    Heightman, E

    T. Heightman, E. Jiang, and A. Acín, Solving the quantum many-body Hamiltonian learning problem with neural differ- ential equations, Quantum Sci. Technol.10, 045072 (2025)

    2025

  57. [57]

    B. C. Hall,Quantum Theory for Mathematicians(Springer, New York, 2013) pp. 333–366. Appendix A: DLA composition Theorem A1.ConsiderKdynamical generating setsA m = {Am,1, Am,2, . . . , Am,Lm },m= 1,2, . . . , K, where each Am generates a dynamical Lie algebrag Am. Letχbe a Her- mitian operator withKdistinct eigenvaluesλ 1, λ2, . . . , λK, and letΠ 1,Π 2, ....

    2013

  58. [58]

    IfF /∈Z(g), then the Lie subalgebraS F ={[F, X]| X∈g}generated by the first-order commutators satisfies ⟨SF ⟩Lie,R =g

    Cyclic DLA Proposition 4.Assumegis a simple Lie algebra overR. IfF /∈Z(g), then the Lie subalgebraS F ={[F, X]| X∈g}generated by the first-order commutators satisfies ⟨SF ⟩Lie,R =g. Proof.Asgis simple, we may invoke the standard root space decomposition relative to a Cartan subalgebrag 0 [36]. For anyX∈g, the root space decomposition gives X=H+ X α cαEα, ...

  59. [59]

    LetF ′ =ϕ(F)

    =g 0. LetF ′ =ϕ(F). SinceF∈g ′ 0, we haveF ′ ∈g 0. Consider the setS F ′ ={[F ′, Y]|Y∈g}. By the former result, we get ⟨SF ′⟩Lie,R =g. 14 Using the property thatϕis a Lie algebra homomorphism pre- serving brackets, we relateS F ′ back toS F , SF ′ ={[ϕ(F), Y]|Y∈g} ={[ϕ(F), ϕ(X)]|X∈g} ={ϕ([F, X])|X∈g} =ϕ(S F ). Therefore, the algebra generated byS F satisf...

  60. [60]

    Non-cyclic DLAs: Derivation of an upper bound for LTFIM and TFIM Lemma A1.[31] ForH=H 0 +αH 1 whereH 1 =Pr γ=1 H γ 1 , the approximate evolution operator Uapx(t) =e −itH0 rY γ=1 eitH0 e−it(H0+αH γ 1 ) approximates the exact evolutionU(t) =e −itH with the error bound ∥U(t)−U apx(t)∥ ≤α2 ˆ t 0 dν ˆ ν 0 ds·(C4) rX γ1<γ2=1 ∥[ ˜H γ1 1 (s), ˜H γ2 1 (ν)]∥,(C5) w...

  61. [61]

    Thus, the sum overγ 1 < γ2 simplifies to a constantC ′, which accounts for the number of non-zero commutator pairs

    Forγ 1 =γ 2, the commutator is trivially zero. Thus, the sum overγ 1 < γ2 simplifies to a constantC ′, which accounts for the number of non-zero commutator pairs. For LTFIM- TFIM,C ′ ≈n 2, since there are n 2 ∼n 2/2pairs ofγ 1 < γ2. Because all commutators∥[ ˜H γ1 1 (s), ˜H γ2 1 (ν)]∥are bounded by a constant, denoted asC ′, then we have ∥ULT F IM(t)−U T ...