Pith. sign in

REVIEW 4 major objections 6 minor 61 references

Sparse conductance measurements can reconstruct disorder-aware Majorana phase maps at one-tenth the measurement cost.

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

2026-08-01 00:28 UTC pith:C76VHTM2

load-bearing objection MEDA is a solid simulation-only result with a real-device claim that the paper doesn't earn; worth refereeing but needs major revision. the 4 major comments →

arxiv 2607.26208 v1 pith:C76VHTM2 submitted 2026-07-28 cs.ET quant-ph

MEDA: Measurement-Efficient Disorder-Aware Majorana Zero Mode Detection in Realistic Devices

Nathan Jones , Binayyak Roy , Valentine Mohaugen , Ian Lewis , Toby Cox , Sumanta Tewari , Rong Ge This is my paper
classification cs.ET quant-ph
keywords Majorana zero modesPeriodic disorder invariantMachine learningConductance spectroscopySparse samplingTopological phase diagramDisordered nanowiresMultiple instance learning
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 the first direct mapping from experimentally measurable conductance data to the periodic disorder invariant (PDI), a bulk-defined, binary topological indicator that stays well-defined in finite, disordered systems. It shows that an attention-based multiple-instance-learning network, trained on simulated disordered nanowire transport, can reconstruct full (chemical potential, Zeeman field) phase diagrams from only 10% of the chemical-potential slices, achieving F1 scores comparable to an idealized scattering-matrix oracle in moderate to strong disorder. If correct, this removes two obstacles to practical Majorana zero mode detection: the inherent bias of boundary-based indicators like the scattering-matrix invariant, and the serial, time-consuming measurement bottleneck of dense parameter sweeps. The paper also asserts that the model's attention weights independently favor conductance features consistent with the topological gap protocol, indicating physical interpretability without explicit protocol-based training.

Core claim

The central claim is that conductance maps, even when sparsely sampled along the chemical potential axis, encode enough disorder-aware bulk information to reproduce the PDI phase diagram of a finite disordered nanowire. The network learns a mapping from a bag of k conductance slices—each containing local and nonlocal differential conductance as a function of bias voltage and Zeeman field—to a full-resolution binary PDI map over (mu, Gamma). This is, to the authors' knowledge, the first demonstration that a bulk-defined, unbiased topological invariant can be inferred from surface observables, and that it can be done with an order-of-magnitude reduction in data acquisition.

What carries the argument

The architecture is an attention-driven multiple-instance-learning (MIL) pipeline. A modified ResNet-18 encoder processes each conductance slice independently to extract local spatial features; absolute coordinate embeddings (mu, V_bias, Gamma) are injected to counter the translation invariance of convolutions, which would otherwise treat a zero-bias peak at V_bias = 0 as equivalent to one at a nonzero bias. A gated attention pooling layer learns to weight each mu-slice by its diagnostic utility, amplifying slices near phase boundaries and suppressing trivial regions, before the encoded slices are aggregated. A CNN decoder then projects the aggregated latent representation into a high-resolu

Load-bearing premise

The central premise is that the quantum transport simulations used for training faithfully reproduce the conductance of real disordered semiconductor-superconductor nanowires, so that a model trained only on synthetic data will map real measured conductance maps to the correct PDI value.

What would settle it

Train the same MEDA architecture on the paper's simulation suite, then run it on experimentally measured differential conductance maps from a real InAs-Al nanowire device for which the PDI has been independently estimated (for example, by reconstructing the Hamiltonian from the known device layout and disorder characterization). If the predicted phase diagram disagrees with the independent PDI estimate at a substantial fraction of (mu, Gamma) points, or if the model shows high confidence in clearly trivial regions, the central claim fails. A simpler partial falsifier would be to feed the model

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

If this is right

  • Experimental phase-diagram mapping in Majorana nanowires could be accelerated by roughly an order of magnitude, since only about 10% of chemical-potential slices need to be measured.
  • The sparse-sampling paradigm could generalize to other parameter axes or to other quantum device characterization tasks where serial parameter sweeps are the primary bottleneck.
  • If the learned attention weights are reliable, they can guide adaptive data acquisition: an experimenter measures the next slice where the model indicates the highest information gain.
  • Because the prediction target is the PDI rather than a boundary-based invariant, predictions should be resilient to quasi-Majorana false positives, giving a cleaner separation between true topological Majorana zero modes and trivial near-zero-energy states.
  • PDI phase maps predicted from sparse conductance data could serve as a screening tool to identify promising devices before committing to expensive full characterization.

Where Pith is reading between the lines

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

  • The paper's reliance on simulated conductance data leaves the real-world deployment claim untested; a natural next step would be a closed-loop experiment where the model directs sparse measurements and its predicted phase map is checked against an independent bulk probe.
  • The attention patterns suggest that only a few mu-slices near phase boundaries carry most of the topological information, so non-uniform, model-guided slice selection could push the measurement reduction well beyond the demonstrated 10x factor.
  • The total-variation regularization, which the paper notes oversimplifies highly fragmented phase landscapes, points to an explicit trade-off: adaptive regularization or boundary-aware losses might recover fragmentation without sacrificing overfitting resistance, an extension the authors leave open.
  • The attention-pooling mechanism could itself be used as a diagnostic to identify which regions of parameter space are most topologically informative, potentially informing nanowire device design.

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

4 major / 6 minor

Summary. The paper proposes MEDA, a machine-learning pipeline that maps sparse, multi-channel differential-conductance measurements (a small number of µ-slices) directly to the periodic disorder invariant (PDI) phase diagram for disordered semiconductor-superconductor nanowires. The input is a bag of k conductance slices in (V_bias, Γ) with four conductance channels; the output is a high-resolution binary PDI map over the full (µ, Γ) grid. The model uses a shared ResNet-18 encoder, coordinate embeddings, gated attention pooling, and a CNN decoder, trained with a composite loss designed to handle class imbalance, phase fragmentation, and missing µ-slices. Training and evaluation data are generated with Kwant, with PDI labels computed from the same underlying Hamiltonian. The paper claims a 10× reduction in measurement volume with F1 performance comparable to an idealized SMI oracle, superiority over a ViT baseline, and attention features consistent with the topological gap protocol.

Significance. If the central mapping were demonstrated to transfer to real devices, the paper would address two genuine bottlenecks: the PDI is defined from the full Hamiltonian and thus is not directly measurable, and dense µ-sweeps are extremely costly in dilution-refrigerator experiments. The in-simulation results are internally plausible, the code is released, and the train/test separation on unseen disorder profiles is a methodological strength. However, the significance is conditional: all evidence is generated by one simulator and one disorder model, so the current contribution is best read as a simulation-level proof of concept. The claimed experimental deployment step is not supported by the evidence presented.

major comments (4)
  1. [Abstract, §III-C, §IV-C] The central deployment claim — 'enabling deployment of a bulk-defined, unbiased topological diagnostic in realistic experimental settings' — is not supported by the evidence. MEDA is trained and evaluated entirely inside one simulator: conductance from Kwant and PDI labels from the same microscopic Hamiltonian and disorder model. The Section IV-C test profiles are 'unseen' only within that generative family (same site-profile model, fixed α, γ, η, L, Δ). This validates self-consistency of the simulator-to-network pipeline, but not transfer to real devices, where temperature, interactions, orbital effects, gate-dependent barriers, and non-uniform disorder distributions are absent from the training distribution. Please add a distribution-shift benchmark (e.g., a second transport code, an altered broadening/temperature model, or a perturbed conductance-generation process) or explicitly refr
  2. [§IV-B, §V-A] The ViT comparison is not controlled. The paper states that ViT [31] was trained on a smaller, coarser parameter space with random l_c ∈ [20,70] nm, whereas MEDA is compared against fixed l_c = 50 nm, and the text acknowledges 'this training mismatch.' In addition, ViT outputs SMI and the conversion to PDI uses Bayes' rule with assumptions about SMI bias errors that are not analyzed. Therefore the claim in §V-A that MEDA 'outperforms ViT' is not established by the presented figure. Either retrain/report a ViT baseline on MEDA's identical data and evaluation protocol, or restrict the text to a qualitative comparison and remove the numeric outperformance claim.
  3. [§IV-A, §III-A2] The F1 and precision numbers are not fully reproducible because the binarization threshold for MEDA's continuous output is not specified. Section III-A2 describes the decoder output as an 'uncalibrated probability' (or logit), but no threshold selection procedure (fixed at 0.5, Youden index, per-disorder tuning, etc.) is given. Since class imbalance is severe and the claimed F1 values are central to the 10× measurement-efficiency result, the threshold must be stated and its sensitivity analyzed.
  4. [§III-B, Table II] 'Realistic device conditions' are demonstrated only within a narrow slice of parameter space. The disorder model has site profiles d, amplitude V0, and correlation length lc, but all other static parameters (α, γ, η, L, Δ) are fixed to single values. No ablation or stress test shows how the learned transport-to-PDI mapping responds to variations in these experimentally variable parameters. The authors should either add robustness experiments across these parameters or temper the 'realistic devices' language to the specific parameter ranges simulated.
minor comments (6)
  1. [Table II vs §IV-D4] The lower bound of V0 is inconsistent: Table II lists V0 ∈ [0.7, 2.5] meV, while §IV-D4 and Figs. 9–11 use V0 ∈ [0.75, 2.5] meV. Please align.
  2. [§III-C, Fig. 4] The PDI convergence threshold referenced in Fig. 4 is not described numerically in the text. Define the rounding/convergence criterion so the ground-truth generation is reproducible.
  3. [§III-D2] The composite loss weights λ1–λ4 are said to be optimized by grid search and included in released code, but the paper itself does not report the values or sensitivity. A short table or statement of the selected weights and the grid range would strengthen the reproducibility of the method.
  4. [§II and §III-C] The PDI is cited to [35] but not defined or even summarized in the text (e.g., the superlattice embedding procedure). A one-sentence operational definition or a key equation would improve self-containedness.
  5. [§V-C, Fig. 8] The interpretability analysis selects only the top 0.01% of attention scores. This is a very small and potentially cherry-picked set. Reporting a distribution of attention-weighted features versus TGP-derived scores across the full test set would provide stronger evidence for the claimed 'autonomous learning' of TGP-like features.
  6. [Various] Minor typographical issues: 'hyperparamter' (§III-D2), 'boundary-induces biases' (§VI), and 'Morteza et al.' in the Fig. 8 caption where the reference list uses the Aghaee et al. author style. Please correct.

Circularity Check

0 steps flagged

No significant circularity: the conductance-to-PDI mapping is a learned inverse problem; real-device transfer is an evidence gap, not a definitional reduction.

full rationale

The claimed derivation chain is: (i) define the PDI ground truth from the nanowire Hamiltonian (following ref. [35], with overlapping authors); (ii) compute conductance observables from the same Hamiltonian with Kwant; (iii) train MEDA to map sparse conductance to the PDI map; (iv) test on unseen disorder profiles from the same generator. This is a supervised inverse problem, not a definitional equivalence: conductance is a nonlinear, low-dimensional projection of the Hamiltonian, and PDI is a different functional of the Hamiltonian, so a network that predicts PDI from conductance must learn a nontrivial map. The holdout test on disorder profiles not seen in training is a genuine generalization check within the simulation. The 10x measurement reduction is the chosen 10-of-100 slice sampling fraction, not a fitted parameter renamed as a prediction; the empirical content is that accuracy is maintained at that fraction. The main self-citation concern—PDI's status as the robust unbiased ground truth is taken from ref. [35] by B. B. Roy and S. Tewari—is mitigated because PDI is computed directly from the Hamiltonian rather than fitted, is also discussed alongside external bias literature (ref. [11]), and is an externally checkable mathematical construction. The absence of real-device or second-simulator validation makes the real-world deployment claims in the Abstract, Introduction, and Conclusion unsupported, but that is a correctness/transfer limitation, not circularity. Accordingly, no circular step meeting the required 'reduction by construction' standard is present.

Axiom & Free-Parameter Ledger

4 free parameters · 5 axioms · 0 invented entities

MEDA introduces no new physical entity; the PDI is prior work from the same group. The main unstated inputs are modeling assumptions about simulation fidelity and transferability, plus tuned ML hyperparameters and the chosen sparse budget.

free parameters (4)
  • Composite loss weights λ1–λ4 = not reported
    Grid-searched hyperparameters balancing BCE, Dice, Focal, and TV losses; they shape the predicted phase-map smoothness and are central to the claimed accuracy.
  • Sparse measurement budget k = 10 (10% of 100 µ-slices)
    Chosen from Fig. 7 as an operational trade-off; not derived from measurement theory. The 10x claim depends directly on this choice.
  • PDI rounding/convergence threshold = not specified in text
    Ground-truth labels are obtained by rounding raw PDI values after a convergence threshold (Fig. 4); the exact threshold is not given, and binary labels depend on it.
  • Model architecture hyperparameters = not fully reported
    ResNet-18 encoder/decoder and attention MIL include many implicit hyperparameters (latent dimension, attention temperature, learning rate, etc.) that are not listed.
axioms (5)
  • domain assumption PDI is a well-defined, binary topological invariant for finite disordered systems (ref [35]).
    Used as ground truth throughout; definition and validity are cited from the same group's prior work and not re-proven here.
  • domain assumption Kwant scattering simulations of differential conductance faithfully represent experimental transport observables in SM-SC nanowires.
    All inputs are synthetic; the paper assumes the four conductance channels (GLL, GRR, GLR, GRL) capture the same physics as real devices.
  • domain assumption Supervised training on simulated data transfers to real devices with unique disorder profiles.
    The central deployment claim rests on this; no experimental data is used to validate the transfer.
  • domain assumption Attention weights are interpretable as physical feature importance, specifically alignment with the topological gap protocol.
    Used to claim physical interpretability; however, top-attention slice selection without a random control makes the claim fragile.
  • domain assumption The SMI oracle computed from the full scattering matrix is an appropriate theoretical ceiling for SMI-based detection.
    The comparison treats SMI as an oracle despite the paper's own critique of SMI bias in disordered systems.

pith-pipeline@v1.3.0-alltime-deepseek · 15723 in / 12053 out tokens · 121312 ms · 2026-08-01T00:28:29.878928+00:00 · methodology

0 comments
read the original abstract

Fault-tolerant topological quantum computing relies on identifying Majorana zero modes (MZMs), but reliable detection in realistic devices remains challenging. Conventional topological indicators are inherently biased in finite, disordered systems, blurring the distinction between true MZMs and trivial states. Furthermore, attempts to map these indicators to real observables via machine learning require dense, expensive conductance measurements, creating a severe scaling bottleneck. To simultaneously address topological bias and measurement limitations, we present MEDA: a Measurement-Efficient, Disorder-Aware framework for MZM detection in realistic devices. MEDA maps sparse, practically obtainable observables directly to the robust periodic disorder invariant (PDI). Using a novel sparse parameter regime, MEDA reduces measurement volume by 10x while maintaining predictive quality, even in moderate to strong disorder regimes that limit conventional methods. Furthermore, MEDA naturally prioritizes input features consistent with the topological gap protocol, demonstrating strong physical interpretability.

Figures

Figures reproduced from arXiv: 2607.26208 by Binayyak Roy, Ian Lewis, Nathan Jones, Rong Ge, Sumanta Tewari, Toby Cox, Valentine Mohaugen.

Figure 1
Figure 1. Figure 1: MEDA Detection Pipeline. Unlike traditional approaches requir￾ing dense sampling and using biased indicators, MEDA achieves scalable, bias-resilient MZM detection by mapping sparse conductance measurements directly to the disorder-aware PDI. • We demonstrate that MEDA accurately reconstructs ex￾tended topological regions using only 10% of the original measurement volume, resulting in massive savings in exp… view at source ↗
Figure 2
Figure 2. Figure 2: Compared to PDI, SMI performs well under weaker disorder (left), [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: MEDA provides a scalable, bias-resilient pipeline for MZM detection [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Top row: Reference raw PDI maps obtained at [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: (a) As V0 increases, MEDA’s µ, Γ parameter window continues to capture topological activity, while ViT’s window becomes trivial. (b) MEDA’s disorder regime is optimized to capture topological activity, while ViT’s regime is mostly trivial. IV. EVALUATION METHODOLOGY A. Evaluation Metrics MEDA’s performance is primarily evaluated using F1 score and precision. The parameter space suffers from severe class im… view at source ↗
Figure 6
Figure 6. Figure 6: Despite using only 10% of observable data, MEDA achieves performance comparable to the SMI Oracle and offers significant improvement over ViT, the state-of-the-art SMI-based prediction pipeline. while improving upon state-of-the-art, thereby establishing MEDA as both theoretically effective and practically compet￾itive. 2) Resolution and Throughput Trade-offs: A central chal￾lenge in experimental deploymen… view at source ↗
Figure 8
Figure 8. Figure 8: MEDA’s trained attention weights heavily prioritize conductance maps with clearly visible conductance features that Morteza et al. establish as [PITH_FULL_IMAGE:figures/full_fig_p009_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: In the low disorder-correlation regime ( [PITH_FULL_IMAGE:figures/full_fig_p009_9.png] view at source ↗
Figure 11
Figure 11. Figure 11: In the high disorder-correlation regime ( [PITH_FULL_IMAGE:figures/full_fig_p010_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: MEDA predicts smooth topological phase transitions and continuous [PITH_FULL_IMAGE:figures/full_fig_p010_12.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 · 12 canonical work pages

  1. [1]

    2n-quasihole states realize2 n−1-dimensional spinor braiding statistics in paired quantum hall states,

    C. Nayak and F. Wilczek, “2n-quasihole states realize2 n−1-dimensional spinor braiding statistics in paired quantum hall states,”Nucl. Phys. B, vol. 479, no. 3, pp. 529–553, 1996

    1996

  2. [2]

    Paired states of fermions in two dimensions with breaking of parity and time-reversal symmetries and the fractional quantum hall effect,

    N. Read and D. Green, “Paired states of fermions in two dimensions with breaking of parity and time-reversal symmetries and the fractional quantum hall effect,”Phys. Rev. B, vol. 61, pp. 10 267–10 297, Apr 2000. [Online]. Available: https://link.aps.org/doi/10.1103/PhysRevB.61.10267

    doi:10.1103/physrevb.61.10267 2000

  3. [3]

    Nonabelions in the fractional quantum Hall effect,

    G. Moore and N. Read, “Nonabelions in the fractional quantum Hall effect,”Nucl. Phys. B, vol. 360, no. 2–3, pp. 362–396, 1991

    1991

  4. [4]

    Non-abelian anyons and topological quantum computation,

    C. Nayak, S. H. Simon, A. Stern, M. Freedman, and S. Das Sarma, “Non-abelian anyons and topological quantum computation,”Rev. Mod. Phys., vol. 80, pp. 1083–1159, Sep 2008. [Online]. Available: https://link.aps.org/doi/10.1103/RevModPhys.80.1083

    doi:10.1103/revmodphys.80.1083 2008

  5. [5]

    Fault-tolerant quantum computation by anyons,

    A. Y . Kitaev, “Fault-tolerant quantum computation by anyons,”Ann. Phys., vol. 303, no. 1, pp. 2–30, 2003

    2003

  6. [6]

    Quantum mechanics of fractional-spin particles,

    F. Wilczek, “Quantum mechanics of fractional-spin particles,”Phys. Rev. Lett., vol. 49, no. 14, p. 957, 1982

    1982

  7. [7]

    Majorana fermions and a topological phase transition in semiconductor-superconductor heterostructures,

    R. M. Lutchyn, J. D. Sau, and S. D. Sarma, “Majorana fermions and a topological phase transition in semiconductor-superconductor heterostructures,”Phys. Rev. Lett., vol. 105, no. 7, p. 077001, 2010

    2010

  8. [8]

    Helical liquids and majorana bound states in quantum wires,

    Y . Oreg, G. Refael, and F. von Oppen, “Helical liquids and majorana bound states in quantum wires,”Phys. Rev. Lett., vol. 105, no. 17, p. 177002, 2010

    2010

  9. [9]

    Non-abelian quantum order in spin-orbit-coupled semiconduc- tors: Search for topological majorana particles in solid-state systems,

    J. D. Sau, S. Tewari, R. M. Lutchyn, T. D. Stanescu, and S. D. Sarma, “Non-abelian quantum order in spin-orbit-coupled semiconduc- tors: Search for topological majorana particles in solid-state systems,” Phys. Rev. B, vol. 82, no. 21, p. 214509, 2010

    2010

  10. [10]

    Generic new platform for topological quantum computation using semiconductor heterostructures,

    J. D. Sau, R. M. Lutchyn, S. Tewari, and S. D. Sarma, “Generic new platform for topological quantum computation using semiconductor heterostructures,”Phys. Rev. Lett., vol. 104, no. 4, p. 040502, 2010

    2010

  11. [11]

    Identifying biases of the Majorana scattering invariant,

    I. A. Day, A. L. R. Manesco, M. Wimmer, and A. R. Akhmerov, “Identifying biases of the Majorana scattering invariant,” SciPost Phys. Core, vol. 8, p. 047, 2025. [Online]. Available: https://scipost.org/10.21468/SciPostPhysCore.8.3.047

    work page doi:10.21468/scipostphyscore.8.3.047 2025

  12. [12]

    Signatures of majorana fermions in hybrid superconductor-semiconductor nanowire devices,

    V . Mourik, K. Zuo, S. M. Frolov, S. Plissard, E. P. Bakkers, and L. P. Kouwenhoven, “Signatures of majorana fermions in hybrid superconductor-semiconductor nanowire devices,”Sci., vol. 336, no. 6084, pp. 1003–1007, 2012

    2012

  13. [13]

    Anomalous zero-bias conductance peak in a nb–insb nanowire–nb hybrid device,

    M. T. Deng, C. L. Yu, G. Y . Huang, M. Larsson, P. Caroff, and H. Q. Xu, “Anomalous zero-bias conductance peak in a nb–insb nanowire–nb hybrid device,”Nano. Lett., vol. 12, no. 12, pp. 6414–6419, Dec 2012. [Online]. Available: https://doi.org/10.1021/nl303758w

    doi:10.1021/nl303758w 2012

  14. [14]

    Zero-bias peaks and splitting in an al–inas nanowire topological superconductor as a signature of majorana fermions,

    A. Das, Y . Ronen, Y . Most, Y . Oreg, M. Heiblum, and H. Shtrikman, “Zero-bias peaks and splitting in an al–inas nanowire topological superconductor as a signature of majorana fermions,”Nat. Phys., vol. 8, no. 12, pp. 887–895, Dec 2012. [Online]. Available: https://doi.org/10.1038/nphys2479

    doi:10.1038/nphys2479 2012

  15. [15]

    The fractional ac josephson effect in a semiconductor–superconductor nanowire as a signature of majorana particles,

    L. P. Rokhinson, X. Liu, and J. K. Furdyna, “The fractional ac josephson effect in a semiconductor–superconductor nanowire as a signature of majorana particles,”Nat. Phys., vol. 8, no. 11, p. 795, 2012

    2012

  16. [16]

    Superconductor-nanowire devices from tunneling to the multichannel regime: Zero-bias oscillations and magnetoconductance crossover,

    H. e. a. Churchill, “Superconductor-nanowire devices from tunneling to the multichannel regime: Zero-bias oscillations and magnetoconductance crossover,”Phys. Rev. B, vol. 87, no. 24, p. 241401, 2013

    2013

  17. [17]

    Anomalous modulation of a zero-bias peak in a hybrid nanowire- superconductor device,

    A. Finck, D. J. Van Harlingen, P. Mohseni, K. Jung, and X. Li, “Anomalous modulation of a zero-bias peak in a hybrid nanowire- superconductor device,”Phys. Rev. Lett., vol. 110, no. 12, p. 126406, 2013

    2013

  18. [18]

    Majorana bound state in a coupled quantum-dot hybrid- nanowire system,

    M. e. a. Deng, “Majorana bound state in a coupled quantum-dot hybrid- nanowire system,”Sci., vol. 354, no. 6319, pp. 1557–1562, 2016

    2016

  19. [19]

    Ballistic superconductivity in semiconductor nanowires,

    H. e. a. Zhang, “Ballistic superconductivity in semiconductor nanowires,”Nat. Comm., vol. 8, p. 16025, 2017

    2017

  20. [20]

    Experimental phase diagram of zero-bias conductance peaks in superconductor/semiconductor nanowire devices,

    J. e. a. Chen, “Experimental phase diagram of zero-bias conductance peaks in superconductor/semiconductor nanowire devices,”Sci. Adv., vol. 3, no. 9, p. e1701476, 2017

    2017

  21. [21]

    Scaling of majorana zero-bias conductance peaks,

    F. e. a. Nichele, “Scaling of majorana zero-bias conductance peaks,” Phys. Rev. Lett., vol. 119, no. 13, p. 136803, 2017

    2017

  22. [22]

    X-shaped and y-shaped andreev resonance profiles in a superconducting quantum dot,

    S. Mi, D. I. Pikulin, M. Marciani, and C. W. J. Beenakker, “X-shaped and y-shaped andreev resonance profiles in a superconducting quantum dot,”J. Exp. Theor. Phys., vol. 119, no. 6, pp. 1018–1027, Dec 2014. [Online]. Available: https://doi.org/10.1134/S1063776114120176

    work page doi:10.1134/s1063776114120176 2014

  23. [23]

    Class d spectral peak in majorana quantum wires,

    D. Bagrets and A. Altland, “Class d spectral peak in majorana quantum wires,”Phys. Rev. Lett., vol. 109, no. 22, p. 227005, 2012

    2012

  24. [24]

    A zero-voltage conductance peak from weak antilocal- ization in a majorana nanowire,

    D. I. Pikulin, J. Dahlhaus, M. Wimmer, H. Schomerus, and C. Beenakker, “A zero-voltage conductance peak from weak antilocal- ization in a majorana nanowire,”New J. Phys., vol. 14, no. 12, p. 125011, 2012

    2012

  25. [25]

    Transport spectroscopy of n s nanowire junctions with majorana fermions,

    E. Prada, P. San-Jose, and R. Aguado, “Transport spectroscopy of n s nanowire junctions with majorana fermions,”Phys. Rev. B, vol. 86, no. 18, p. 180503, 2012

    2012

  26. [26]

    Physical mechanisms for zero-bias conduc- tance peaks in majorana nanowires,

    H. Pan and S. D. Sarma, “Physical mechanisms for zero-bias conduc- tance peaks in majorana nanowires,”Phys. Rev. Res., vol. 2, no. 1, p. 013377, 2020

    2020

  27. [27]

    Two-terminal charge tun- neling: Disentangling majorana zero modes from partially separated andreev bound states in semiconductor-superconductor heterostructures,

    C. Moore, T. D. Stanescu, and S. Tewari, “Two-terminal charge tun- neling: Disentangling majorana zero modes from partially separated andreev bound states in semiconductor-superconductor heterostructures,” Phys. Rev. B, vol. 97, no. 16, p. 165302, 2018

    2018

  28. [28]

    Quantized zero-bias conductance plateau in semiconductor-superconductor heterostructures without topological majorana zero modes,

    C. Moore, C. Zeng, T. D. Stanescu, and S. Tewari, “Quantized zero-bias conductance plateau in semiconductor-superconductor heterostructures without topological majorana zero modes,”Phys. Rev. B, vol. 98, p. 155314, Oct 2018. [Online]. Available: https://link.aps.org/doi/10.1103/PhysRevB.98.155314

    work page doi:10.1103/physrevb.98.155314 2018

  29. [29]

    Reproducing topological properties with quasi-Majorana states,

    A. Vuik, B. Nijholt, A. R. Akhmerov, and M. Wimmer, “Reproducing topological properties with quasi-Majorana states,”SciPost Phys., vol. 7, p. 061, 2019

    2019

  30. [30]

    Robust low-energy andreev bound states in semiconductor-superconductor structures: Importance of partial separation of component majorana bound states,

    T. D. Stanescu and S. Tewari, “Robust low-energy andreev bound states in semiconductor-superconductor structures: Importance of partial separation of component majorana bound states,”Phys. Rev. B, vol. 100, no. 15, p. 155429, 2019

    2019

  31. [31]

    Vision transformer based deep learning of topological indicators in majorana nanowires,

    J. R. Taylor and S. Das Sarma, “Vision transformer based deep learning of topological indicators in majorana nanowires,”Phys. Rev. B, vol. 111, p. 104208, Mar 2025. [Online]. Available: https://link.aps.org/doi/10.1103/PhysRevB.111.104208

    work page doi:10.1103/physrevb.111.104208 2025

  32. [32]

    Quantized conductance at the majorana phase transition in a disordered superconducting wire,

    A. R. Akhmerov, J. P. Dahlhaus, F. Hassler, M. Wimmer, and C. W. J. Beenakker, “Quantized conductance at the majorana phase transition in a disordered superconducting wire,”Phys. Rev. Lett., vol. 106, p. 057001, 2011

    2011

  33. [33]

    Scattering theory of topological insulators and superconductors,

    I. C. Fulga, F. Hassler, and A. R. Akhmerov, “Scattering theory of topological insulators and superconductors,”Phys. Rev. B, vol. 85, p. 165409, 2012

    2012

  34. [34]

    How to infer non-abelian statistics and topological visibility from tunneling conductance properties of realistic majorana nanowires,

    S. Das Sarma, A. Nag, and J. D. Sau, “How to infer non-abelian statistics and topological visibility from tunneling conductance properties of realistic majorana nanowires,”Phys. Rev. B, vol. 94, p. 035143, 2016

    2016

  35. [35]

    Topological invariant for finite systems in the presence of disorder,

    R. Eissele, B. B. Roy, S. Tewari, and T. D. Stanescu, “Topological invariant for finite systems in the presence of disorder,” 2025. [Online]. Available: https://arxiv.org/abs/2508.13146

    Pith/arXiv arXiv 2025

  36. [36]

    Machine learning the disorder landscape of majorana nanowires,

    J. R. Taylor, J. D. Sau, and S. Das Sarma, “Machine learning the disorder landscape of majorana nanowires,”Phys. Rev. Lett., vol. 132, p. 206602, May 2024. [Online]. Available: https://link.aps.org/doi/10.1103/PhysRevLett.132.206602

    work page doi:10.1103/physrevlett.132.206602 2024

  37. [37]

    Machine learning detection of majorana zero modes from zero-bias peak measurements,

    M. Cheng, R. Okabe, A. Chotrattanapituk, and M. Li, “Machine learning detection of majorana zero modes from zero-bias peak measurements,” Matter, vol. 7, no. 7, pp. 2507–2520, 2024. [Online]. Available: https://www.sciencedirect.com/science/article/pii/S2590238524002534

    2024

  38. [38]

    Ballistic majorana nanowire devices,

    ¨O. e. a. G ¨ul, “Ballistic majorana nanowire devices,”Nat. Nano., vol. 13, no. 3, pp. 192–197, 2018

    2018

  39. [39]

    Exponential protection of zero modes in majorana islands,

    S. M. e. a. Albrecht, “Exponential protection of zero modes in majorana islands,”Nat., vol. 531, no. 7593, pp. 206–209, 2016

    2016

  40. [40]

    Inas-al hybrid devices passing the topological gap protocol,

    M. e. a. Aghaee, “Inas-al hybrid devices passing the topological gap protocol,”Phys. Rev. B, vol. 107, p. 245423, Jun 2023. [Online]. Available: https://link.aps.org/doi/10.1103/PhysRevB.107.245423

    doi:10.1103/physrevb.107.245423 2023

  41. [41]

    Computer-automated tuning of semiconductor double quantum dots into the single-electron regime,

    T. A. Baart, P. T. Eendebak, C. Reichl, W. Wegscheider, and L. M. K. Vandersypen, “Computer-automated tuning of semiconductor double quantum dots into the single-electron regime,”Appl. Phys. Lett., vol. 108, no. 21, May 2016. [Online]. Available: http://dx.doi.org/10.1063/1.4952624

    doi:10.1063/1.4952624 2016

  42. [42]

    Autonomous tuning and charge-state detection of gate- defined quantum dots,

    J. e. a. Darulov ´a, “Autonomous tuning and charge-state detection of gate- defined quantum dots,”Phys. Rev. Appl., vol. 13, no. 5, May 2020. [On- line]. Available: http://dx.doi.org/10.1103/PhysRevApplied.13.054005

    doi:10.1103/physrevapplied.13.054005 2020

  43. [43]

    Transport signatures of quasiparticle poisoning in a majorana island,

    S. e. a. Albrecht, “Transport signatures of quasiparticle poisoning in a majorana island,”Phys. Rev. Lett., vol. 118, no. 13, p. 137701, 2017

    2017

  44. [44]

    Hybridization of subgap states in one-dimensional superconductor-semiconductor coulomb islands,

    E. e. a. O Farrell, “Hybridization of subgap states in one-dimensional superconductor-semiconductor coulomb islands,”Phys. Rev. Lett., vol. 121, no. 25, p. 256803, 2018

    2018

  45. [45]

    Parity transitions in the superconducting ground state of hybrid insb–al coulomb islands,

    J. e. a. Shen, “Parity transitions in the superconducting ground state of hybrid insb–al coulomb islands,”Nat. Comm., vol. 9, no. 1, p. 4801, 2018

    2018

  46. [46]

    Normal, superconducting and topological regimes of hybrid double quantum dots,

    D. Sherman, J. Yodh, S. M. Albrecht, J. Nyg ˚ard, P. Krogstrup, and C. M. Marcus, “Normal, superconducting and topological regimes of hybrid double quantum dots,”Nat. Nano., vol. 12, no. 3, p. 212, 2017

    2017

  47. [47]

    Selective-area-grown semiconductor- superconductor hybrids: A basis for topological networks,

    S. e. a. Vaitiek ˙enas, “Selective-area-grown semiconductor- superconductor hybrids: A basis for topological networks,”Phys. Rev. Lett., vol. 121, no. 14, p. 147701, 2018

    2018

  48. [48]

    Non-majorana states yield nearly quantized conductance in proximatized nanowires,

    P. e. a. Yu, “Non-majorana states yield nearly quantized conductance in proximatized nanowires,”Nat. Phys., vol. 17, no. 4, p. 482–488, Jan

  49. [49]

    Large zero-bias peaks in insb-al hybrid semiconductor- superconductor nanowire devices,

    H. Z. et al., “Large zero-bias peaks in insb-al hybrid semiconductor- superconductor nanowire devices,” 2021

    2021

  50. [50]

    Near-zero-energy end states in topologically trivial spin-orbit coupled superconducting nanowires with a smooth confinement,

    G. Kells, D. Meidan, and P. Brouwer, “Near-zero-energy end states in topologically trivial spin-orbit coupled superconducting nanowires with a smooth confinement,”Phys. Rev. B, vol. 86, no. 10, p. 100503, 2012

    2012

  51. [51]

    Stabilizing topological superconductivity in disordered spin- orbit coupled semiconductor-superconductor heterostructures,

    B. B. Roy, R. Jaiswal, T. D. Stanescu, and S. Tewari, “Stabilizing topological superconductivity in disordered spin- orbit coupled semiconductor-superconductor heterostructures,”Phys. Rev. B, vol. 110, p. 115436, Sep 2024. [Online]. Available: https://link.aps.org/doi/10.1103/PhysRevB.110.115436

    work page doi:10.1103/physrevb.110.115436 2024

  52. [52]

    Majorana fermions in semiconductor nanowires: fundamentals, modeling, and experiment,

    T. D. Stanescu and S. Tewari, “Majorana fermions in semiconductor nanowires: fundamentals, modeling, and experiment,”J. Phys. Cond. Matt., vol. 25, no. 23, p. 233201, May 2013. [Online]. Available: http://dx.doi.org/10.1088/0953-8984/25/23/233201

    work page doi:10.1088/0953-8984/25/23/233201 2013

  53. [53]

    To close or not to close: The fate of the superconducting gap across the topological quantum phase transition in majorana-carrying semiconductor nanowires,

    T. D. Stanescu, S. Tewari, J. D. Sau, and S. Das Sarma, “To close or not to close: The fate of the superconducting gap across the topological quantum phase transition in majorana-carrying semiconductor nanowires,”Phys. Rev. Lett., vol. 109, p. 266402, Dec 2012. [Online]. Available: https://link.aps.org/doi/10.1103/PhysRevLett.109.266402

    work page doi:10.1103/physrevlett.109.266402 2012

  54. [54]

    Scattering formula for the topological quantum number of a disordered multimode wire,

    I. C. Fulga, F. Hassler, A. R. Akhmerov, and C. W. J. Beenakker, “Scattering formula for the topological quantum number of a disordered multimode wire,”Phys. Rev. B, vol. 83, p. 155429, Apr 2011. [Online]. Available: https://link.aps.org/doi/10.1103/PhysRevB.83.155429

    work page doi:10.1103/physrevb.83.155429 2011

  55. [55]

    Protocol to identify a topological superconducting phase in a three-terminal device,

    D. I. P. et al., “Protocol to identify a topological superconducting phase in a three-terminal device,” 2021. [Online]. Available: https://arxiv.org/abs/2103.12217

    Pith/arXiv arXiv 2021

  56. [56]

    Reproducing topological properties with quasi-majorana states,

    A. Vuik, B. Nijholt, A. R. Akhmerov, and M. Wimmer, “Reproducing topological properties with quasi-majorana states,”arXiv, 2018

    2018

  57. [57]

    Parametric exploration of zero-energy modes in three-terminal insb-al nanowire devices,

    J.-Y . Wang, N. van Loo, G. P. Mazur, V . Levajac, F. K. Malinowski, M. Lemang, F. Borsoi, G. Badawy, S. Gazibegovic, E. P. A. M. Bakkers, M. Quintero-P ´erez, S. Heedt, and L. P. Kouwenhoven, “Parametric exploration of zero-energy modes in three-terminal insb-al nanowire devices,”Phys. Rev. B, vol. 106, p. 075306, Aug 2022. [Online]. Available: https://l...

    work page doi:10.1103/physrevb.106.075306 2022

  58. [58]

    Density of states of disordered topological superconductor-semiconductor hybrid nanowires,

    J. D. Sau and S. Das Sarma, “Density of states of disordered topological superconductor-semiconductor hybrid nanowires,”Phys. Rev. B, vol. 88, no. 6, Aug. 2013. [Online]. Available: http://dx.doi.org/10.1103/PhysRevB.88.064506

    work page doi:10.1103/physrevb.88.064506 2013

  59. [59]

    Kwant: a software package for quantum transport,

    C. W. Groth, M. Wimmer, A. R. Akhmerov, and X. Waintal, “Kwant: a software package for quantum transport,”New J. Phys., vol. 16, no. 6, p. 063065, Jun. 2014. [Online]. Available: http://dx.doi.org/10.1088/1367- 2630/16/6/063065

    doi:10.1088/1367- 2014

  60. [60]

    Quantized majorana conductance,

    H. Z. et al., “Quantized majorana conductance,”Nat., vol. 556, pp. 74– 79, 2018

    2018

  61. [2021]

    Available: http://dx.doi.org/10.1038/s41567-020-01107- w

    [Online]. Available: http://dx.doi.org/10.1038/s41567-020-01107- w

    work page doi:10.1038/s41567-020-01107-