arxiv: 2604.15269 · v1 · submitted 2026-04-16 · 🪐 quant-ph · cs.LG· math.ST· stat.TH

Recognition: unknown

Cloning is as Hard as Learning for Stabilizer States

Nikhil Bansal , Matthias C. Caro , Gaurav Mahajan

Authors on Pith no claims yet

Pith reviewed 2026-05-10 11:22 UTC · model grok-4.3

classification 🪐 quant-ph cs.LGmath.STstat.TH

keywords stabilizer statesquantum cloningsample complexityquantum learningno-cloning theoremsample amplificationquantum cryptography

0 comments

The pith

For n-qubit stabilizer states the optimal cloning sample complexity is Θ(n), so cloning remains as hard as learning.

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

The paper examines how the no-cloning principle interacts with structure in quantum states. It focuses on n-qubit stabilizer states and proves that cloning them requires a linear number of copies in n. This matches the number of copies needed to learn the state. The finding shows that structure does not reduce the difficulty of cloning relative to learning for this class. The work therefore gives a finer view of no-cloning that links quantum foundations to learning theory.

Core claim

For n-qubit stabilizer states, the optimal sample complexity of cloning is Θ(n). Thus, also for this structured class of states, cloning is as hard as learning.

What carries the argument

A reduction from stabilizer-state cloning to a structured sample-amplification problem on linear distributions.

If this is right

Cloning lower bounds are obtained by proving new sample-amplification lower bounds for classes of distributions with linear structure.
The no-cloning theorem admits a fine-grained version that applies to stabilizer states.
Connections are established between quantum foundations, learning theory, and quantum cryptography.

Where Pith is reading between the lines

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

Similar hardness results may hold for other structured families such as product states or states with limited entanglement.
Cryptographic protocols that assume easy copying of certain states may need to be re-examined when the states are stabilizers.
The linear lower bound supplies a concrete benchmark against which future approximate-cloning schemes for structured states can be tested.

Load-bearing premise

The reduction from stabilizer-state cloning to the structured sample-amplification problem on linear distributions is valid and tight.

What would settle it

Either an explicit cloning protocol that succeeds with o(n) copies of an unknown stabilizer state or a direct proof that the corresponding sample-amplification lower bound for linear distributions fails.

read the original abstract

The impossibility of simultaneously cloning non-orthogonal states lies at the foundations of quantum theory. Even when allowing for approximation errors, cloning an arbitrary unknown pure state requires as many initial copies as needed to fully learn the state. Rather than arbitrary unknown states, modern quantum learning theory often considers structured classes of states and exploits such structure to develop learning algorithms that outperform general-state tomography. This raises the question: How do the sample complexities of learning and cloning relate for such structured classes? We answer this question for an important class of states. Namely, for $n$-qubit stabilizer states, we show that the optimal sample complexity of cloning is $\Theta(n)$. Thus, also for this structured class of states, cloning is as hard as learning. To prove these results, we use representation-theoretic tools in the recently proposed Abelian State Hidden Subgroup framework and a new structured version of the recently introduced random purification channel to relate stabilizer state cloning to a variant of the sample amplification problem for probability distributions that was recently introduced in classical learning theory. This allows us to obtain our cloning lower bounds by proving new sample amplification lower bounds for classes of distributions with an underlying linear structure. Our results provide a more fine-grained perspective on No-Cloning theorems, opening up connections from foundations to quantum learning theory and quantum cryptography.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

Stabilizer-state cloning requires Θ(n) copies and matches learning complexity via a reduction to classical sample amplification on linear distributions.

read the letter

The main takeaway is that for n-qubit stabilizer states the optimal number of copies needed to clone is Θ(n), the same order as learning them. The authors reach this by mapping the cloning task to a structured sample-amplification problem on linear distributions, using the Abelian State Hidden Subgroup framework plus a new structured random purification channel they define. This produces matching lower bounds that line up with the known Ω(n) learning lower bound for the same class. The reduction looks independent of the quantum claim and does not rely on fitted parameters inside the paper. The representation-theoretic steps and the channel construction appear internally consistent, with the linear structure preserved as claimed. What the paper does cleanly is give a fine-grained no-cloning statement for an important structured family rather than arbitrary states, and it opens an explicit link between quantum foundations and classical learning theory. The tightness of the bound is a plus; they do not overclaim generality. A minor soft spot is that the new channel is purpose-built, so a reader has to check the definition and error analysis carefully, though the supplied details show no gaps. The work is fairly technical and will be easiest for people already comfortable with quantum representation theory or quantum learning. It is aimed at researchers who want to see how structure affects sample complexity in cloning versus learning, or who work on related questions in quantum cryptography. I would bring the paper to a reading group to discuss the reduction technique. It deserves a serious referee because the central claim is sharp, the method is new in this setting, and the evidence checks out without circularity. I would recommend sending it to peer review rather than desk-rejecting.

Referee Report

0 major / 3 minor

Summary. The paper proves that the optimal sample complexity of approximately cloning n-qubit stabilizer states is Θ(n), matching the known Ω(n) lower bound for learning the same class. The argument reduces stabilizer cloning to a structured sample-amplification problem over linear distributions by combining the Abelian State Hidden Subgroup framework with a newly introduced structured random purification channel, then establishes matching classical lower bounds for the resulting distribution class.

Significance. If the central reduction holds, the result supplies a fine-grained no-cloning statement for an important structured family, showing that the stabilizer structure does not decouple cloning from learning. The technical contributions—the structured channel and the new sample-amplification lower bounds for linear distributions—are likely to be reusable in quantum learning and cryptography. The manuscript earns credit for deriving the lower bound via an independent classical argument rather than by circular appeal to quantum parameters.

minor comments (3)

§3.2: the definition of the structured random purification channel is given abstractly; adding a short explicit matrix representation for n=2 would help readers verify that it indeed maps stabilizer states to linear distributions.
Theorem 5.1 and the surrounding text: the notation for the error parameter ε is overloaded between the quantum cloning fidelity and the classical amplification distance; a single clarifying sentence would remove ambiguity.
References: the citation to the original random-purification channel (Ref. [X]) appears only in the introduction; repeating the pointer in §4 when the structured variant is defined would improve traceability.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for their positive summary of our work and for recommending minor revision. The report correctly captures the main result—that approximate cloning of n-qubit stabilizer states requires Θ(n) samples, matching the learning lower bound—and the technical approach via the Abelian State Hidden Subgroup framework and the structured random purification channel. No specific major comments were provided in the report, so we have no individual points to rebut. We are happy to make any minor editorial or presentational changes requested by the editor.

Circularity Check

0 steps flagged

No significant circularity; derivation self-contained

full rationale

The paper establishes the Θ(n) cloning lower bound for n-qubit stabilizer states by reducing the quantum problem to a classical sample-amplification lower bound on linear distributions. This reduction is performed via the Abelian State Hidden Subgroup framework together with a newly introduced structured random purification channel; both the reduction and the classical lower bounds are derived within the manuscript and do not rely on fitted parameters or prior self-citations for their validity. The matching upper bound follows directly from the known Ω(n) learning complexity for the same class, which is external to the cloning argument. No self-definitional steps, renamed predictions, or load-bearing self-citation chains appear in the core derivation chain.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 1 invented entities

The central claim rests on the validity of the Abelian State Hidden Subgroup framework for stabilizer states and on the correctness of the newly defined structured random purification channel that reduces cloning to sample amplification.

axioms (2)

domain assumption Representation-theoretic tools in the Abelian State Hidden Subgroup framework apply to stabilizer states
Invoked to relate cloning to hidden-subgroup problems
ad hoc to paper The structured random purification channel preserves the relevant linear structure of stabilizer states
New construction introduced in the paper to obtain the reduction

invented entities (1)

structured random purification channel no independent evidence
purpose: Relate stabilizer cloning to classical sample amplification on linearly structured distributions
New object defined to obtain the lower-bound reduction; no independent evidence outside the paper is provided in the abstract

pith-pipeline@v0.9.0 · 5537 in / 1312 out tokens · 30054 ms · 2026-05-10T11:22:55.969882+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

27 extracted references · 25 canonical work pages

[1]

Reed-Muller Codes for Random Era- sures and Errors

[ASW15] Emmanuel Abbe, Amir Shpilka, and Avi Wigderson. “Reed-Muller Codes for Random Era- sures and Errors”. In:Proceedings of the Forty-Seventh Annual ACM Symposium on Theory of Computing. STOC ’15. Portland, Oregon, USA: Association for Computing Machinery, 2015, pp. 297–306.isbn: 9781450335362.doi: 10.1145/2746539.2746575.url: https://doi.org/10.1145/...

work page doi:10.1145/2746539.2746575.url: 2015
[2]

Testing and Learning Structured Quantum Hamiltonians

Leibniz International Proceedings in Informatics (LIPIcs). Dagstuhl, Germany: Schloss Dagstuhl – Leibniz-Zentrum für Informatik, 2023, 3:1–3:24.isbn: 978-3-95977-283-9. doi: 10.4230/LIPIcs.TQC.2023.3.url: https://drops.dagstuhl.de/entities/document/10.4230/ LIPIcs.TQC.2023.3 (pages 3, 10). [ADE25] Srinivasan Arunachalam, Arkopal Dutt, and Francisco Escude...

work page doi:10.4230/lipics.tqc.2023.3.url: 2023
[3]

On Approximability of Satisfiable k-CSPs: IV , year =

Vancouver, BC, Canada: Association for Computing Ma- chinery, 2024, pp. 1470–1477.isbn: 9798400703836.doi: 10.1145/3618260.3649619.url: https: //doi.org/10.1145/3618260.3649619 (page 5). [BDK16] Charles H. Baldwin, Ivan H. Deutsch, and Amir Kalev. “Strictly-complete measurements for bounded-rank quantum-state tomography”. In:Phys. Rev. A93 (5 May 2016), p...

work page doi:10.1145/3618260.3649619.url: 2024
[4]

Bittel, J

arXiv: 2504.12263[quant-ph].url: https: //arxiv.org/abs/2504.12263 (page 5). [Blu+25] Andreas Bluhm, Matthias C. Caro, Francisco Escudero Gutiérrez, Aadil Oufkir, and Cam- byse Rouzé.Certifying and learning quantum Ising Hamiltonians

work page arXiv
[5]

Stabilizer bootstrapping: A recipe for efficient agnostic tomography and magic estimation

arXiv: 2509 . 10239 [quant-ph].url: https://arxiv.org/abs/2509.10239 (page 5). [BNZ25] John Bostanci, Barak Nehoran, and Mark Zhandry. “A General Quantum Duality for Represen- tations of Groups with Applications to Quantum Money, Lightning, and Fire”. In:Proceedings of the 57th Annual ACM Symposium on Theory of Computing. STOC ’25. Prague, Czechia: As- so...

work page doi:10.1145/3717823.3718195 2025
[6]

Dagstuhl, Germany: Schloss Dagstuhl – Leibniz-Zentrum für Informatik, 2024, 24:1–24:23.isbn: 978-3-95977-309-6.doi: 10.4230/LIPIcs.ITCS.2024.24

Leibniz International Proceedings in Informatics (LIPIcs). Dagstuhl, Germany: Schloss Dagstuhl – Leibniz-Zentrum für Informatik, 2024, 24:1–24:23.isbn: 978-3-95977-309-6.doi: 10.4230/LIPIcs.ITCS.2024.24. url: https://drops.dagstuhl.de/entities/document/10.4230/LIPIcs.ITCS.2024.24 (page 3). 44 [CNS25] Matthias C. Caro, Preksha Naik, and Joseph Slote.Testin...

work page doi:10.4230/lipics.itcs.2024.24 2024
[7]

[CAN25] Chi-Fang Chen, Anurag Anshu, and Quynh T

arXiv: 2411.12730[quant-ph].url: https://arxiv.org/abs/2411.12730 (page 3). [CAN25] Chi-Fang Chen, Anurag Anshu, and Quynh T. Nguyen.Learning quantum Gibbs states locally and efficiently

work page arXiv
[8]

Plenio and Steven T

arXiv: 2504.02706[quant-ph].url: https://arxiv.org/abs/2504.02706 (page 5). [Cra+10] Marcus Cramer, Martin B Plenio, Steven T Flammia, Rolando Somma, David Gross, Stephen D Bartlett, Olivier Landon-Cardinal, David Poulin, and Yi-Kai Liu. “Efficient quantum state tomography”. In:Nature communications1.1 (2010), p. 149.doi: 10.1038/ncomms1147 (pages 3, 5). ...

work page doi:10.1038/ncomms1147 2010
[9]

15343[quant-ph]

arXiv: 2504 . 15343[quant-ph]. url: https://arxiv.org/abs/2504.15343 (pages 3, 10). [Fin03] Steven R Finch.Mathematical constants. Cambridge university press, 2003.url: https : / / www.cambridge.org/gb/universitypress/subjects/mathematics/recreational- mathematics/ mathematical-constants-1?format=PB (page 23). [FH13] William Fulton and Joe Harris.Represen...

work page arXiv 2003
[10]

In: Staffa, M.,Cabibihan,J.J.,Siciliano,B.,Ge,S.S.,Bodenhagen,L.,Tapus,A.,Rossi,S.,Cav- allo, F., Fiorini, L., Matarese, M., He, H

Springer Science & Business Media, 2013.url: https://link.springer.com/book/10.1007/978- 1- 4612- 0979- 9 (page 10). [GML25] Filippo Girardi, Francesco Anna Mele, and Ludovico Lami.Random purification channel made simple

work page doi:10.1007/978- 2013
[11]

Girardi, F

arXiv: 2511.23451[quant-ph].url: https://arxiv.org/abs/2511.23451 (pages 15, 31). [Gir+25] Filippo Girardi, Francesco Anna Mele, Haimeng Zhao, Marco Fanizza, and Ludovico Lami. Random Stinespring superchannel: converting channel queries into dilation isometry queries

work page arXiv
[12]

On the Complexity of Teaching

arXiv: 2512.20599[quant-ph].url: https://arxiv.org/abs/2512.20599 (page 15). [GK95] S.A. Goldman and M.J. Kearns. “On the Complexity of Teaching”. In:Journal of Computer and System Sciences50.1 (1995), pp. 20–31.issn: 0022-0000.doi: https://doi.org/10.1006/jcss.1995. 1003.url: https://www.sciencedirect.com/science/article/pii/S0022000085710033 (pages 6, 9...

work page doi:10.1006/jcss.1995 1995
[13]

Quantum State Tomography via Compressed Sensing

issn: 2521-327X.doi: 10.22331/q-2025-11-06-1907.url: http://dx.doi.org/10.22331/q-2025- 11-06-1907 (page 5). [Gro+10] David Gross, Yi-Kai Liu, Steven T. Flammia, Stephen Becker, and Jens Eisert. “Quantum State Tomography via Compressed Sensing”. In:Phys. Rev. Lett.105 (15 Oct. 2010), p. 150401.doi: 10.1103/PhysRevLett.105.150401.url: https://link.aps.org/...

work page doi:10.22331/q-2025-11-06-1907.url: 2025
[14]

Optimal Online Discrepancy Minimization

arXiv: 2505.15770[quant-ph].url: https://arxiv. org/abs/2505.15770 (pages 7, 12, 13). [Hua+24] Hsin-Yuan Huang, Yunchao Liu, Michael Broughton, Isaac Kim, Anurag Anshu, Zeph Lan- dau, and Jarrod R. McClean. “Learning Shallow Quantum Circuits”. In:Proceedings of the 56th Annual ACM Symposium on Theory of Computing. STOC ’24. ACM, June 2024, pp. 1343–1351. ...

work page doi:10.1145/3618260.3649722.url: 2024
[15]

Learning t-doped stabilizer states

arXiv: 1002 . 4632[quant-ph].url: https : / / arxiv. org / abs / 1002 . 4632 (pages 3, 5). [LOH24] Lorenzo Leone, Salvatore F. E. Oliviero, and Alioscia Hamma. “Learning t-doped stabilizer states”. In:Quantum8 (May 2024), p. 1361.issn: 2521-327X.doi: 10.22331/q-2024-05-27-1361. url: https://doi.org/10.22331/q-2024-05-27-1361 (page 5). [Li+22] Gene Li, Pri...

work page doi:10.22331/q-2024-05-27-1361 2024
[16]

23737–23750.url: https : / / proceedings

Curran Associates, Inc., 2022, pp. 23737–23750.url: https : / / proceedings . neurips . cc / paper _ files / paper / 2022 / file / 960cfbb846aff424ac20aadce6fa6530-Paper-Conference.pdf (page 9). 46 [Mel25] Antonio Anna Mele.Lecture Notes on Representation Theory for Quantum Information (QMATH Masterclass 2025, Copenhagen). Lecture Notes. 2025.url: https :...

2022
[17]

Optimal learning of quantum channels in diamond distance

arXiv: 2512.10214[quant-ph].url: https://arxiv.org/abs/2512.10214 (page 15). [Mel+25] Francesco Anna Mele, Filippo Girardi, Senrui Chen, Marco Fanizza, and Ludovico Lami.Ran- dom purification channel for passive Gaussian bosons

work page arXiv
[18]

Optimal Counterfeiting Attacks and Gen- eralizations for Wiesner’s Quantum Money

arXiv: 2512 . 16878[quant-ph]. url: https://arxiv.org/abs/2512.16878 (page 15). [MVW13] Abel Molina, Thomas Vidick, and John Watrous. “Optimal Counterfeiting Attacks and Gen- eralizations for Wiesner’s Quantum Money”. In:Theory of Quantum Computation, Commu- nication, and Cryptography. Ed. by Kazuo Iwama, Yasuhito Kawano, and Mio Murao. Berlin, Heidelberg...

work page arXiv 2013
[19]

Learning stabilizer states by Bell sampling

arXiv: 1707 . 04012 [quant-ph].url: https://arxiv.org/abs/1707.04012 (pages 3, 5, 6, 8). [NZ24] Barak Nehoran and Mark Zhandry. “A Computational Separation Between Quantum No- Cloning and No-Telegraphing”. In:15th Innovations in Theoretical Computer Science Confer- ence (ITCS 2024). Ed. by Venkatesan Guruswami. Vol

work page Pith review arXiv 2024
[20]

Efficient quantum tomography

Leibniz International Proceedings in Informatics (LIPIcs). Dagstuhl, Germany: Schloss Dagstuhl – Leibniz-Zentrum für Infor- matik, 2024, 82:1–82:23.isbn: 978-3-95977-309-6.doi: 10 . 4230 / LIPIcs . ITCS . 2024 . 82.url: https://drops.dagstuhl.de/entities/document/10.4230/LIPIcs.ITCS.2024.82 (pages 3, 10). [NC10] Michael A Nielsen and Isaac L Chuang.Quantu...

work page doi:10.4230/lipics.itcs.2024.82 2024
[21]

In:arXiv:2511.15806 [quant-ph](2025)

arXiv: 2511.15806[quant-ph].url: https://arxiv.org/abs/2511. 15806 (page 15). [PSW25] Angelos Pelecanos, Jack Spilecki, and John Wright.The debiased Keyl’s algorithm: a new un- biased estimator for full state tomography

work page arXiv
[22]

Quantum hypergraph states

arXiv: 2510.07788[quant-ph].url: https: //arxiv.org/abs/2510.07788 (pages 3, 5). [Ros+13] M Rossi, M Huber, D Bruß, and C Macchiavello. “Quantum hypergraph states”. In:New Journal of Physics15.11 (Nov. 2013), p. 113022.issn: 1367-2630.doi: 10.1088/1367-2630/15/11/113022. url: http://dx.doi.org/10.1088/1367-2630/15/11/113022 (page 10). [RS24] Cambyse Rouzé...

work page doi:10.1088/1367-2630/15/11/113022 2013
[23]

1977 , publisher =

Springer, 1977.url: https: //link.springer.com/book/10.1007/978-1-4684-9458-7 (page 10). [Sho94] P.W. Shor. “Algorithms for quantum computation: discrete logarithms and factoring”. In:Pro- ceedings 35th Annual Symposium on Foundations of Computer Science. 1994, pp. 124–134.doi: 10.1109/SFCS.1994.365700 (page 13). [Slo64] N. J. A. Sloane.A048651, OEIS Foun...

work page doi:10.1007/978-1-4684-9458-7 1977
[24]

Valiant , title =

07622[quant-ph].url: https://arxiv.org/abs/2510.07622 (pages 7–9, 15, 31). [Val84] L. G. Valiant. “A theory of the learnable”. In:Commun. ACM27.11 (Nov. 1984), pp. 1134–1142. issn: 0001-0782.doi: 10.1145/1968.1972.url: https://doi.org/10.1145/1968.1972 (page 4). [VC71] V. N. Vapnik and A. Ya. Chervonenkis. “On the Uniform Convergence of Relative Frequen- ...

work page doi:10.1145/1968.1972.url: 1984
[25]

PMLR, 30 Jun– 04 Jul 2025, pp

Proceedings of Machine Learning Research. PMLR, 30 Jun– 04 Jul 2025, pp. 5553–5604.url: https://proceedings.mlr.press/v291/vasconcelos25a.html (page 5). [WW25] Michael Walter and Freek Witteveen.A random purification channel for arbitrary symmetries with applications to fermions and bosons

2025
[26]

Optimal cloning of pure states

arXiv: 2512 . 15690[quant-ph].url: https : //arxiv.org/abs/2512.15690 (pages 9, 15, 33). [Wer98] R. F. Werner. “Optimal cloning of pure states”. In:Physical Review A58.3 (Sept. 1998), pp. 1827– 1832.issn: 1094-1622.doi: 10 . 1103 / physreva . 58 . 1827.url: http : / / dx . doi . org / 10 . 1103 / PhysRevA.58.1827 (pages 3, 5, 7, 10). [Wie83] Stephen Wiesn...

work page doi:10.1145/1008908.1008920.url: 1998
[27]

Learning Quantum States and Unitaries of Bounded Gate Complexity

arXiv: 2512.21260[quant-ph].url: https://arxiv.org/abs/2512.21260 (page 15). [Zha+24] Haimeng Zhao, Laura Lewis, Ishaan Kannan, Yihui Quek, Hsin-Yuan Huang, and Matthias C. Caro. “Learning Quantum States and Unitaries of Bounded Gate Complexity”. In:PRX Quan- tum5 (4 Oct. 2024), p. 040306.doi: 10.1103/PRXQuantum.5.040306.url: https://link.aps.org/ doi/10....

work page doi:10.1103/prxquantum.5.040306.url: 2024