REVIEW 3 major objections 4 minor 1 cited by
Low-Fidelity Entanglement Distillation with FIMAX
T0 review · 3 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash
Pith's one-line read FIMAX, a stabilizer-based two-copy protocol, distills low-fidelity entangled states that defeat other two-copy methods.
desk verdict FIMAX's low-fidelity advantage is a real numerical finding, but the paper overstates it as a universal theorem in Sec. 3.3. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing object is FIMAX's choice of a stabilizer subgroup of the two-copy Weyl error group acting on Alice's side. For each stabilizer, generated by a single Weyl operator $W(g)$, the protocol partitions two-copy error elements first by their symplectic product with $g$ into sets $\varepsilon(s)$, and then into cosets of the stabilizer. FIMAX picks the stabilizer and coset $(C_{\max}, s_{\max})$ maximizing the probability ratio $P(C_{\max})/P(\varepsilon(s_{\max}))$, performs stabilizer measurements on both copies, keeps the event where the outcome difference equals $s_{\max}$, and applies the inverse canonic encoding followed by a Weyl correction. This makes the dominant error class a success, so the output state has higher fidelity with the target maximally entangled state.
What would settle it
Find a single low-fidelity NPT state in $d=3$ with $\langle\Omega_{k,l}|\rho_{\mathrm{in}}|\Omega_{k,l}\rangle \le 1/3$ for all $(k,l)$ that ADGJ can distill but FIMAX cannot; the paper's claims that every such state distillable by another protocol is also FIMAX-distillable, and that FIMAX is the only protocol with a significant share, would then fail.
Extended reading notes
Core claim
The central claim is that FIMAX outperforms all other considered recurrent two-copy protocols on low-fidelity NPT states, and the gap widens as the fidelity restriction tightens. In the paper's numerical samples with fidelity below $1/d$, FIMAX distills 43% of random pure NPT states for $d=2$ and 54% for $d=3$, whereas the next-best protocol, ADGJ, manages 32% and 2%; for standard Bell-diagonal states the corresponding shares are 100% and 92%, against 77% and 3%. With the stronger restriction $\langle\Omega_{k,l}|\rho_{\mathrm{in}}|\Omega_{k,l}\rangle \le 1/d$ for every Bell state, no considered protocol distills anything in $d=2$, while in $d=3$ FIMAX still distills 33% of pure states and 83% of Bell-diagonal states, with ADGJ below 1%. The authors interpret this as evidence that the stabilizer structure exploited by FIMAX lets it convert noise patterns that defeat other recurrence protocols into successful distillation events.
Load-bearing premise
The numerical comparison assumes that the sampled 10,000 pure and Bell-diagonal states per dimension, and 100 diagonal states for each of 1,000 generalized Bell bases, are representative of all low-fidelity NPT states in those families, and that the software implementing FIMAX matches the protocol of Ref. [13].
Editorial extensions
If this is right
- For $d=3$ under the strict fidelity bound, FIMAX is the only considered protocol with a non-negligible distillable share, so stabilizing a quantum link in this noise regime may require a stabilizer-based protocol rather than existing recurrence schemes.
- Because every FIMAX-distillable state reaches fidelity at least $1/d$ after one iteration, FIMAX can serve as a first-stage purifier whose output is handed to higher-threshold protocols.
- Because every FIMAX-undistillable NPT state becomes PPT after one iteration, a single FIMAX step decides two-copy distillability for the sampled families, saving repetitions.
- The inclusion observation, that everything other protocols distill FIMAX also distills, implies FIMAX's success set contains the union of the other protocols' success sets, at least within the numerical samples.
Reading between the lines
- If the inclusion observation is exact rather than a sampling artifact, then FIMAX's stabilizer optimization may define a largest distillable class for two-copy recurrence protocols, and the share of distillable states becomes a property of the state family rather than of the chosen protocol.
- The one-step characterization suggests a practical preprocessing test for quantum repeaters: apply a single FIMAX round and check whether the output is PPT; this is cheaper than running the full recurrence to convergence.
- The numerical data cover only prime $d=2$ and $d=3$; a natural extension is to test whether the hierarchy persists for higher prime dimensions and whether composite dimensions, where stabilizer generators are no longer single, change the picture.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript reports a numerical comparison of the stabilizer-based FIMAX protocol against ADGJ, DEJMPS, P1-P2, and generalized BBPSSW for two-copy entanglement distillation of low-fidelity bipartite states in dimensions d=2 and d=3. The authors sample pure states, standard Bell-diagonal states, and generalized Bell-diagonal states; impose NPT and low-fidelity restrictions; and compare the share of states each protocol can distill. They find that FIMAX distills the largest share in every state family, with a particularly large advantage for d=3 under the strict restriction that all Bell fidelities are at most 1/d. The paper also lists qualitative observations about universal dominance of FIMAX, one-iteration PPT failure, and one-iteration fidelity thresholds, which are subsequently restated as conclusions.
Significance. If the reported shares are correct, the paper provides useful evidence that FIMAX is more robust to noise than other recurrent two-copy protocols in the tested state families, especially for qutrits. The quantitative comparison is against independent benchmark protocols and is not circular; the use of an open-source Julia package and the explicit protocol description in Sec. 5 support reproducibility. The practical significance is moderate: this is a numerical performance benchmark rather than an analytic characterization of FIMAX's distillable region. The headline share numbers are plausible, but the stronger universal claims in Sec. 3.3 go beyond the finite data and need to be either restricted to sampled states or proven.
major comments (3)
- [Sec. 3 and Sec. 5.3.1] The paper does not define the operational criterion by which a state is counted as 'distillable' for the reported shares. In particular, it is not stated whether a state counts if one iteration gives any increase in fidelity, if the output fidelity exceeds 1/d, if the output is NPT, or if repeated application converges to a maximally entangled state. This matters because Sec. 3.3 frames the results in terms of one-iteration behavior and because recurrent protocols are usually assessed by asymptotic convergence. Please state the criterion explicitly and, if only one iteration is used, justify it as a measure of distillability.
- [Sec. 3.3 and Sec. 4] The universal statements in Sec. 3.3 — 'all states that can be distilled with a different protocol, can also be distilled with FIMAX' and 'all NPT states that cannot be distilled with FIMAX, result in a PPT output state' — are supported only by finite random samples: 10,000 pure and Bell-diagonal states and 100 states per 1,000 generalized Bell bases. The stated ≤1% deviation concerns aggregate share estimates, not universal 'for all' statements, and the paper itself notes the limited accuracy of the generalized Bell-diagonal max and min values. Sec. 4 then restates these observations as established facts. Please either restrict all such claims to 'all sampled states' or provide a proof or exact characterization of FIMAX's success region.
- [Sec. 3.1 and Figs. 1-2] For non-Bell-diagonal pure states, FIMAX's first step applies the Weyl twirl, while the comparison protocols are not described as doing so. It is therefore unclear whether the pure-state shares in Figs. 1 and 2 are computed on the same input states for all protocols or whether FIMAX benefits from or suffers from the twirl. Please specify the exact state fed into each protocol and how probabilistic success events are handled, since this directly affects the reported pure-state shares.
minor comments (4)
- [Sec. 3.2] For d=2 under the strict fidelity restriction, it is not clear whether any NPT states satisfying the restriction exist at all; if the set is empty, the statement that none of the protocols can distill any state is vacuous. Please clarify whether the share is defined over an empty set or whether NPT states were actually found.
- [Sec. 3.2] The protocol name 'P12' appears in the sentence beginning 'For FIMAX, BBPSSW and P12...'; this should be 'P1-P2' to match the notation used elsewhere.
- [Sec. 5.3.2] In the example, 'correspond to the larges value' should read 'correspond to the largest value'.
- [Sec. 5.3.1] The protocol description would benefit from an explicit definition of the notation P(C) and P(ε(s)) in terms of the two-copy density matrix elements introduced in Eq. (8), since the current text uses these probabilities before fully specifying their relation to ρ⊗2.
Circularity Check
No significant circularity: FIMAX is evaluated against independent benchmark protocols; its performance is measured externally, not baked into its definition.
full rationale
The paper's central claim is a comparative numerical analysis of the FIMAX protocol against ADGJ, DEJMPS, P1-P2, and BBPSSW. FIMAX's decision rule (maximizing the coset-to-sector probability ratio) is fixed by the authors' prior work Ref. [13] and is applied state-by-state to random samples; the reported distillable shares are computed from output fidelities, an external success criterion, not from the optimization objective itself. No parameter is fitted to the data and then renamed a prediction. The comparison protocols are independent external benchmarks, so the superiority claim has content beyond the construction of FIMAX. Self-citations to Ref. [13] and Ref. [19] supply the protocol definition and the code being tested; they do not by themselves certify the empirical outcome. The universal statements in Sec. 3.3 are extrapolated from finite samples, which is a sampling/correctness concern rather than circularity: the paper itself notes the limited accuracy of the 100-state generalized-Bell samples for max/min values. Because the performance measurement is not equivalent to the protocol's input by construction, no circular step is exhibited.
Assumptions & free parameters
assumptions (4)
- domain assumption Any stabilizer relevant for entanglement distillation in prime dimensions is generated by a single error operator W(g).
- standard math The Weyl twirl channel (11) maps any state to Bell-diagonal form and is LOCC-realizable.
- domain assumption Any bipartite state with fidelity > 1/d is distillable by some recurrent two-copy protocol.
- standard math PPT entangled states are bound entangled and therefore undistillable.
Cite this review
Pith. "Pith review of Low-Fidelity Entanglement Distillation with FIMAX." pith.science (2026). https://pith.science/paper/VFBHCMD7
@misc{pith2026250209261,
author = {Pith},
title = {Pith review of: Low-Fidelity Entanglement Distillation with FIMAX},
year = {2026},
howpublished = {\url{https://pith.science/paper/VFBHCMD7}},
note = {Machine review of arXiv:2502.09261}
}
abstract
Uncontrolled interactions with the environment introduce errors that remain a significant challenge to the reliability of quantum technologies using entanglement. An essential method to overcome or mitigate these errors is entanglement distillation, the transformation of multiple copies of weakly entangled states into a smaller number of approximately maximally entangled states. We present a comparative analysis of the stabilizer-based two-copy entanglement distillation protocol, FIMAX, against other recurrent two-copy protocols, including ADGJ, DEJMPS, P1-P2, and the generalized BBPSSW protocol. We focus on low-fidelity bipartite quantum states in dimensions $d = 2$ and $d = 3$, which are particularly challenging to distill. Our findings demonstrate that FIMAX exhibits superior performance for these states. While other protocols struggle with highly noisy states, FIMAX successfully distills entanglement even when the initial state quality is severely compromised. These results highlight the protocol's capability to address the effects of environmental noise, advancing the robustness and scalability of quantum technologies leveraging entanglement distillation.
Figures
Forward citations
Cited by 1 Pith paper
-
Quantification of the energy consumption of entanglement distribution
The paper proves that entanglement irreversibility implies a non-zero lower bound on the standard energy cost of distributing an ebit through a noisy quantum channel.
Reference graph
Works this paper leans on
-
[13]
Bennett, C. H. et al. Teleporting an unknown quan- tum state via dual classical and Einstein-Podolsky- Rosen channels. Physical Review Letters 70. Pub- lisher: American Physical Society, 1895–1899.https: //link.aps.org/doi/10.1103/PhysRevLett.70. 1895 (2024) (Mar. 1993)
-
[1]
Pirandola, S. et al. Advances in quantum cryp- tography. EN. Advances in Optics and Photonics
- [2]
-
[3]
Chitambar, E. & Gour, G. Quantum resource the- ories. Reviews of Modern Physics 91. Publisher: American Physical Society, 025001. https://link. aps.org/doi/10.1103/RevModPhys.91.025001 (2025) (Apr. 2019)
-
[4]
B¨ auml, S., Das, S., Wang, X. & Wilde, M. M. Re- source theory of entanglement for bipartite quan- tum channels. arXiv:1907.04181 [quant-ph]. http: //arxiv.org/abs/1907.04181 (2024) (July 2019)
arXiv 2024
-
[5]
Popp, C. & Hiesmayr, B. C. Special features of the Weyl–Heisenberg Bell basis imply unusual entan- glement structure of Bell-diagonal states. en. New Journal of Physics 26. Publisher: IOP Publishing, 013039. issn: 1367-2630. https://dx.doi.org/10. 1088/1367-2630/ad1d0e (2024) (Jan. 2024)
work page 2024
-
[6]
Quantum Computing in the NISQ era and beyond
Preskill, J. Quantum Computing in the NISQ era and beyond. en-GB. Quantum 2. Publisher: Verein zur F¨ orderung des Open Access Publizierens in den Quantenwissenschaften, 79. https://quantum- journal.org/papers/q-2018-08-06-79/ (2025) (Aug. 2018)
work page 2025
-
[7]
Acharya, R. et al. Quantum error correction below the surface code threshold. en. Nature. Publisher: Nature Publishing Group, 1–7. issn: 1476-4687. https : / / www . nature . com / articles / s41586 - 024-08449-y (2025) (Dec. 2024)
work page 2025
Show all 36 references
-
[8]
Bravyi, S. et al. High-threshold and low-overhead fault-tolerant quantum memory. en. Nature 627. Publisher: Nature Publishing Group, 778–782. issn: 1476-4687. https://www.nature.com/articles/ s41586-024-07107-7 (2025) (Mar. 2024)
2025
-
[9]
Steane, A. M. Error Correcting Codes in Quan- tum Theory. Physical Review Letters 77. Publisher: American Physical Society, 793–797. https://link. aps . org / doi / 10 . 1103 / PhysRevLett . 77 . 793 (2025) (July 1996)
2025
-
[11]
& Briegel, H
D¨ ur, W. & Briegel, H. J. Entanglement purifica- tion and quantum error correction. en. Reports on Progress in Physics 70, 1381. issn: 0034-4885. https://dx.doi.org/10.1088/0034-4885/70/8/ R03 (2024) (July 2007)
2024 doi
-
[12]
issn: 1943-8206
Publisher: Optica Publishing Group, 1012–1236. issn: 1943-8206. https://opg.optica.org/aop/ abstract.cfm?uri=aop- 12- 4- 1012(2024) (Dec. 2020)
2024
-
[14]
Popp, C., Sutter, T. C. & Hiesmayr, B. C. A Novel Stabilizer-based Entanglement Distillation Proto- col for Qudits. arXiv:2408.02383 [quant-ph]. http: //arxiv.org/abs/2408.02383 (2025) (Aug. 2024)
2025
-
[15]
Stabilizer Codes and Quantum Er- ror Correction
Gottesman, D. Stabilizer Codes and Quantum Er- ror Correction. en. arXiv:quant-ph/9705052. http: //arxiv.org/abs/quant-ph/9705052 (2024) (May 1997)
2024 arXiv
-
[16]
Conversion of a general quantum sta- bilizer code to an entanglement distillation protocol
Matsumoto, R. Conversion of a general quantum sta- bilizer code to an entanglement distillation protocol. en. Journal of Physics A: Mathematical and General
-
[17]
& Horodecki, R
Horodecki, M., Horodecki, P. & Horodecki, R. Mixed- State Entanglement and Distillation: Is there a “Bound” Entanglement in Nature? en. Physical Re- view Letters 80, 5239–5242. issn: 0031-9007, 1079-
-
[18]
Vollbrecht, K. G. H. & Wolf, M. M. Efficient dis- tillation beyond qubits. en. Physical Review A 67, 012303. issn: 1050-2947, 1094-1622. https://link. aps . org / doi / 10 . 1103 / PhysRevA . 67 . 012303 (2024) (Jan. 2003). 9
2024
-
[19]
BellDiagonalQudits: A package for entan- glement analyses of mixed maximally entangled qudits
Popp, C. BellDiagonalQudits: A package for entan- glement analyses of mixed maximally entangled qudits. en. Journal of Open Source Software 8,
-
[20]
& Horodecki, P
Horodecki, M. & Horodecki, P. Reduction criterion of separability and limits for a class of distillation protocols. Physical Review A 59. Publisher: Amer- ican Physical Society, 4206–4216. https://link. aps.org/doi/10.1103/PhysRevA.59.4206 (2024) (June 1999)
2024 doi
-
[21]
C., Popp, C
Hiesmayr, B. C., Popp, C. & Sutter, T. C. Bipartite Bound Entanglement. arXiv:2406.13491 [quant-ph]. http://arxiv.org/abs/2406.13491 (2024) (June 2024)
2024 arXiv
-
[22]
& Jex, I
Alber, G., Delgado, A., Gisin, N. & Jex, I. Efficient bipartite quantum state purification in arbitrary di- mensional Hilbert spaces. en. Journal of Physics A: Mathematical and General 34. arXiv:quant- ph/0102035, 8821–8833. issn: 0305-4470, 1361-6447. http : / / arxiv . org /...
2024
-
[23]
Deutsch, D. et al. Quantum Privacy Amplification and the Security of Quantum Cryptography over Noisy Channels. Physical Review Letters 77. Pub- lisher: American Physical Society, 2818–2821.https: //link.aps.org/doi/10.1103/PhysRevLett.77. 2818 (2024) (Sept. 1996)
2024 doi
-
[24]
& D¨ ur, W
Miguel-Ramiro, J. & D¨ ur, W. Efficient entanglement purification protocols for d -level systems. en. Physi- cal Review A 98, 042309. issn: 2469-9926, 2469-9934. https://link.aps.org/doi/10.1103/PhysRevA. 98.042309 (2024) (Oct. 2018)
2024 doi
-
[25]
& Lewen- stein, M
˙Zyczkowski, K., Horodecki, P., Sanpera, A. & Lewen- stein, M. Volume of the set of separable states. en. Physical Review A 58, 883–892. issn: 1050-2947, 1094-1622. https://link.aps.org/doi/10.1103/ PhysRevA.58.883 (2024) (Aug. 1998)
2024
-
[26]
Non-binary unitary error bases and quan- tum codes en
Knill, E. Non-binary unitary error bases and quan- tum codes en. Tech. rep. LA-UR–96-2717, 373768 (June 1996), LA–UR–96–2717, 373768. http : / / www.osti.gov/servlets/purl/373768- BfsVyz/ webviewable/ (2024)
2024
-
[27]
Hiesmayr, B. C. & L¨ offler, W. Complementarity re- veals bound entanglement of two twisted photons. en. New Journal of Physics 15, 083036. issn: 1367-
-
[28]
& Narnhofer, H
Baumgartner, B., Hiesmayr, B. & Narnhofer, H. A special simplex in the state space for entangled qu- dits. en. Journal of Physics A: Mathematical and Theoretical 40. arXiv:quant-ph/0610100, 7919–7938. issn: 1751-8113, 1751-8121. http://arxiv.org/ abs/quant-ph/0610100 (2024) (J...
2024 arXiv
-
[29]
Bennett, C. H. et al. Purification of Noisy Entangle- ment and Faithful Teleportation via Noisy Channels. en. Physical Review Letters 76, 722–725. issn: 0031- 9007, 1079-7114. https://link.aps.org/doi/10. 1103/PhysRevLett.76.722 (2024) (Jan. 1996)
2024
-
[30]
& Hiesmayr, B
Popp, C. & Hiesmayr, B. C. Comparing bound entan- glement of bell diagonal pairs of qutrits and ququarts. en. Scientific Reports 13. Number: 1 Publisher: Nature Publishing Group, 2037. issn: 2045-2322. https : / / www . nature . com / articles / s41598 - 023-29211-w (2024) (Feb. 2023)
2024
-
[31]
Nonbinary quantum codes
Rains, E. Nonbinary quantum codes. IEEE Transac- tions on Information Theory 45. Conference Name: IEEE Transactions on Information Theory, 1827–
-
[34]
& Hiesmayr, B
Popp, C. & Hiesmayr, B. C. Almost complete solu- tion for the NP-hard separability problem of Bell diagonal qutrits. en. Scientific Reports 12. Number: 1 Publisher: Nature Publishing Group, 12472. issn: 2045-2322. https://www.nature.com/articles/ s41598-022-16225-z (2024) (July 2022)
2024
-
[36]
issn: 0305- 4470, 1361-6447
arXiv:quant-ph/0209091, 8113–8127. issn: 0305- 4470, 1361-6447. http://arxiv.org/abs/quant- ph/0209091 (2024) (July 2003)
2024 arXiv
-
[1832]
BellDiagonalQudits.jl
issn: 1557-9654. https://ieeexplore.ieee. org/document/782103 (2024) (Sept. 1999). Acknowledgments C.P. and B.C.H. acknowledge gratefully that this re- search was funded in whole, or in part, by the Aus- trian Science Fund (FWF) project P36102-N (Grant DOI: 10 10.55776/P36102)...
2024 doi
-
[2630]
1088/1367-2630/15/8/083036 (2024) (Aug
https://iopscience.iop.org/article/10. 1088/1367-2630/15/8/083036 (2024) (Aug. 2013)
2024
-
[4924]
https://joss.theoj.org/ papers/10.21105/joss.04924 (2024) (Jan
issn: 2475-9066. https://joss.theoj.org/ papers/10.21105/joss.04924 (2024) (Jan. 2023)
2024 doi
-
[7114]
https : / / link . aps . org / doi / 10 . 1103 / PhysRevLett.80.5239 (2024) (June 1998)
2024
Reviewed August 7, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.