Pith. sign in

REVIEW 3 major objections 5 minor

Quantum random-number generator with non-demolition measurements: semi-device-independent implementation

T0 review · 3 major / 5 minor · reviewed 2026-07-30 · grok-4.5

Pith's one-line read A tripartite non-demolition setup certifies quantum path superposition on one detector while a second detector emits near-uniform three-outcome random numbers at the same time, without needing spacelike separation.

desk verdict Solid dual-detector QNDM QRNG idea with clean algebra and a real min-entropy trade-off; the semi-DI/source-independent security claim overreaches what the model actually proves. read the letter →

arxiv 2607.27025 v2 pith:KNPU3SFN submitted 2026-07-29 quant-ph

classification quant-ph
keywords quantumrandom-numbergeneratornon-demolitionmeasurementsemi-device-independentquasi-probabilitydistributionpathsuperpositionmin-entropyLeggett-Garg
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper proposes a quantum random-number generator built from one three-level system coupled to two detectors (two-level and three-level) under sequential quantum non-demolition measurements. One detector reconstructs a quasi-probability distribution over evolution paths; negativity in that distribution certifies coherent path superposition and therefore genuine quantum dynamics. The other detector simultaneously produces outcomes from a three-value alphabet whose average distribution can be tuned near uniform by choice of initial state and unitary, so the min-entropy of the output string stays close to the maximum. Generation and certification happen in the same runs, and certification does not require spacelike separation of devices. That removes a major barrier of Bell-based device-independent generators and points toward compact, integrable hardware while still giving a semi-device-independent guarantee against source manipulation, provided the detector measurements themselves are trusted.

What carries the argument

The quantum non-demolition measurement quasi-probability distribution P_ND(Δ) = P_cl + P_q reconstructed from D1: negativity of P_q is treated as a necessary and sufficient signature of path superposition that a classical probability source cannot reproduce, while the same couplings set the three-outcome statistics of D2.

What would settle it

Build the proposed three-level system plus two detectors, reconstruct P_ND from D1 while collecting D2 strings, and check whether clear negativity appears together with a near-uniform three-outcome average and min-entropy consistent with the stated bound; absence of negativity when the designed unitary and state should produce it, or classical spoofing of the same negativity under the actual couplings, would refute the claim.

Watch

Extended reading notes

Core claim

In a QNDM tripartite architecture, detector D1’s reconstructed quasi-probability P_ND certifies path superposition via negativity while detector D2 simultaneously emits outcomes β in {0, ±2} whose λ-averaged distribution can be optimized near-uniform. The min-entropy of the full string is lower-bounded by LK log(3/(1+3|P_q(−1)|)), and an optimized spin-1 example reaches about 90% of maximal extractable randomness while still showing clear negativity.

Load-bearing premise

The security claim rests on trusting the measurements of both detectors and on the assertion that observed negativity in the quasi-probability cannot be faked by a classical device sampling a genuine probability distribution.

Editorial extensions

If this is right

  • Randomness generation and quantum certification can run on the same compact chip without spacelike-separated modules.
  • Initial state and unitary can be tuned to trade a controlled amount of min-entropy for stronger negativity certification.
  • Adding further non-demolition couplings or higher-dimensional detectors is predicted to enlarge the outcome alphabet and improve extractable randomness.
  • The protocol supplies a concrete semi-device-independent (source-independent) QRNG route that needs no input random seed.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Because certification and generation share the same physical runs, finite-sample statistical tests of negativity directly bound the usable randomness rate in real time.
  • Integrated platforms that already support spin-1 or qutrit control (cold atoms, superconducting circuits, NV centers) are natural first testbeds for a proof-of-principle device.
  • If classical models can reproduce the observed P_ND negativity under realistic noise and finite L, the semi-DI security argument would need an explicit noise-tolerant reformulation.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The paper proposes a QRNG based on a tripartite quantum non-demolition measurement setup: a three-level system S coupled to a two-level detector D1 and a three-level detector D2. D1 reconstructs a quasi-probability P_ND whose negativity certifies path superposition, while D2 simultaneously emits outcomes β∈{0,±2} whose λ-averaged distribution can be optimized near uniformity. A min-entropy bound H_min(X)≥LK log(3/(1+3|P_q(−1)|)) links extractable randomness to certified negativity. An optimized spin-1 example achieves relative min-entropy deficit ~9.81% (bound δ≈15.2% at |P_q(−1)|≈0.06). Certification needs no spacelike separation. The protocol is presented as semi-device-independent/source-independent with trusted detector measurements.

Significance. If the construction and scoped security claims hold, this is a practically relevant alternative to Bell and Leggett–Garg QRNGs: simultaneous generation and certification without spatial separation, plus an explicit trade-off between quasi-probability negativity and min-entropy. The unitary evolution, reduced detector states, path indexing, Â degeneracy, and P_q(β)↔P_q(Δ) identities are derived carefully in the text and Appendix. The optimization example is concrete. Even with a narrower security model, the architecture contributes to semi-DI randomness and to applications of QNDM witnesses.

major comments (3)
  1. [Random number generator; Eq. (5); Appendix (S17)–(S31)] The source-independent/semi-DI claim overreaches what is proven. The text says the user need not know preparations “not even … assumptions on the state dimensions,” and that Eq. (5) certifies randomness against source manipulation. The bound and identities P_q(β=0)=cos(λ_l)P_q(−1), P_q(β=2)=−cos(λ_l)P_q(−1) (Appendix after S22–S24) hold only inside the fixed model: 3-level S, stated â/Â spectra, U1/U2 couplings, and path degeneracy tying A_4,7 to β=0. No reduction shows an arbitrary source reproducing the observed G_λ must obey the same D2 bias. Restate security to match trusted readouts and model assumptions, or supply a proper reduction.
  2. [System and set-up; Eqs. (3)–(4); P_ND decomposition] Security against classical spoofing of P_ND negativity under this protocol is imported from prior QNDM work [21,22] without a self-contained argument for these couplings and finite (L,K) sampling. The claim that negative regions cannot be reproduced by a classical device sampling a probability distribution must address D2’s decoherence functional f(n,k) and finite Fourier sampling of G_λ. A short classical-simulation bound under trusted readouts is needed for certification to be load-bearing here.
  3. [Random number generator; Conclusions] The trusted vs untrusted partition is under-specified for a semi-DI claim. The paper notes that “measurements of detectors D1 and D2 need to be trusted,” yet still frames the protocol as source-independent in the sense of Refs. [24,25]. Clarify in one place which operations are trusted (D1/D2 measurements, couplings, dimensions) and which are untrusted (source state and û), and what adversarial power observed negativity alone rules out.
minor comments (5)
  1. [Fig. 2] Figure 2 caption/panel labels are hard to parse in the manuscript rendering. Ensure the published figure clearly separates P_ND(Δ) from P̄_β and reports the numerical parameters.
  2. [System and set-up; Appendix] The mapping n=1+3(i+1)+(j+1) is dense; a small table of (i,j)↔n, α_n, β_n would help verify degeneracy β∈{0,±2} and the identification of A_4,7, A_5,8, A_6,9.
  3. [Application; Conclusions] In Application/Conclusions, “more than 90% … genuinely random” should be tied explicitly to ΔH_min/H̄_min∼9.81% versus the looser δ≈15.2% bound.
  4. [Introduction; Eqs. (1)–(2)] Typos: “a posteriorion” → “a posteriori on”; occasional ρ_f vs ρ̄_f switches; M in U2 is not fixed numerically in the example.
  5. [Introduction; Conclusions] Brief comparison of rate and entropy per shot with LGI QRNGs [15,16] and photonic source-independent QRNGs would better position the practical gain.

Circularity Check

1 steps flagged · score 2.0 of 10

Mild self-citation load on the QNDM negativity criterion; min-entropy trade-off is derived inside the model, not forced by definition.

  1. self citation load bearing [Introduction; System and set-up (after Eq. 3 / P_ND decomposition); Refs. [21, 22]]
    "it has been shown that QNDM provides a stronger criterion than the LGI for the identification and certification of quantum behavior [21, 22]. ... the negativity of P_ND(Δ) provides a necessary and sufficient signature of path superposition in the dynamics [21, 22]. Such negative regions cannot be reproduced by a classical device that generates the corresponding strings from an underlying probability distribution."

    The operational meaning of D1 certification—that observed negativity of the reconstructed quasi-probability is necessary and sufficient for genuine path superposition and cannot be faked by a classical probability sampler—is not derived in this manuscript; it is taken from the authors’ earlier QNDM papers. The QRNG security/certification narrative rests on that imported lemma. This is load-bearing self-citation for the certification claim, though the entropy algebra itself is computed independently inside the present model.

full rationale

The paper’s central algebraic content—the path-labeled detectors’ state, the explicit link P_q(β=0)=cos(λ_l)P_q(−1) and P_q(β=2)=−cos(λ_l)P_q(−1), and the resulting min-entropy lower bound H_min(X)≥LK log(3/(1+3|P_q(−1)|))—is derived in the main text and Appendix from the assumed tripartite couplings, spectra, and degeneracy structure. That is a legitimate trade-off calculation within a fixed model, not a quantity redefined as its own input. The only circularity-adjacent element is interpretive: the claim that negativity of P_ND is a necessary and sufficient signature of path superposition that a classical sampler cannot reproduce is imported from the authors’ prior QNDM papers [21, 22] rather than re-proved here. That self-citation is load-bearing for the certification narrative and the semi-DI framing, but it does not make the entropy bound or the optimized near-uniform example tautological. No fitted-input-called-prediction, uniqueness-import, or ansatz-smuggling pattern appears. Score 2 reflects one mild self-citation burden with independent central derivation.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The paper sits on standard finite-dimensional QM plus the authors’ prior QNDM quasi-probability certification criterion. Load-bearing modeling choices are commuting path observables, fixed small Hilbert-space dimensions, trusted detector measurements, and hand-optimized pure input state and J_x rotation. No new physical entities are postulated; free parameters are the preparation/evolution knobs used to illustrate near-uniformity.

free parameters (3)
  • Initial state amplitudes ψ0_i and relative phase Δφ=φ1−φ0 = ψ0_−1≈0.76, |ψ0_0|≈|ψ0_1|≈0.46, Δφ=π/2
    Chosen/optimized to maximize |P_q(−1)| subject to uniform classical margins P_cl(β)=1/3; example values ψ0_−1≈0.76, |ψ0_0|≈|ψ0_1|≈0.46, Δφ=π/2.
  • Evolution angle θ in û=exp(iθ Ĵx) = θ≈π/3
    Tied to |ψ_−1| by the uniform-P_cl constraint and set near π/3 in the example to maximize certification-relevant negativity.
  • Discretization L and shots K per λ_l = L≳30 (example); K free
    Protocol parameters controlling reconstruction of G_λ and string length; large-L average used to drop oscillatory terms; example claims L≳30 for the quoted entropy deviation.
assumptions (5)
  • standard math Standard quantum mechanics on finite-dimensional Hilbert spaces with unitary couplings and Born-rule readout of detectors.
    Used throughout for evolution (S1)–(S3) and outcome probabilities.
  • domain assumption â and  on S commute and share the stated eigenbasis (a_i=i, A_i=1 for i=0,1 and A_−1=−1), enabling joint path labeling.
    Stated after the coupling unitaries; required for the nine-path picture and degeneracy that lets D2 outcomes interfere on D1.
  • domain assumption Negativity of the QNDM quasi-probability P_ND is necessary and sufficient for path superposition and cannot be reproduced by a classical device sampling a genuine probability distribution.
    Imported from Refs. [21,22] and used as the certification criterion for D1.
  • domain assumption Measurements on D1 and D2 are trusted; only the source/evolution need not be characterized (semi-DI / source-independent stance).
    Explicitly stated in the Random number generator section; without it the protocol is not a certificate against malicious devices.
  • ad hoc to paper Detector dimensions and coupling forms (phase encoding U1, U2 with M-level Fourier readout on D2) are as assumed; user need not know preparations but dimensions are fixed in the analysis.
    The concrete 2+3+3 level choice and Â=J_z^2 degeneracy structure are design choices that make three-outcome near-uniform randomness work.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Quantum random-number generator with non-demolition measurements: semi-device-independent implementation." pith.science (2026). https://pith.science/paper/KNPU3SFN

@misc{pith2026260727025,
  author       = {Pith},
  title        = {Pith review of: Quantum random-number generator with non-demolition measurements: semi-device-independent implementation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/KNPU3SFN}},
  note         = {Machine review of arXiv:2607.27025}
}
read the original abstract

We propose and analyze a novel quantum random-number generator based on a tripartite quantum system in which two subsystems act as detectors. Within a quantum non-demolition measurement scheme, one detector is used to certify the presence of genuine quantum effects in the system's evolution, while the second generates random numbers from a distribution that can be optimized to maximize their entropy. Using one two-level system and two three-level systems, we generate random numbers from a nearly uniform three-outcome distribution, yielding close-to-maximal entropy and therefore near-optimal randomness generation. A key feature of the protocol is that randomness generation and certification occur simultaneously. Moreover, certification does not rely on spacelike separation between detectors, removing a major constraint of device-independent approaches. This property enables practical implementation and facilitates the miniaturization of the device, making the protocol a promising candidate for scalable quantum technologies.

Figures

Figures reproduced from arXiv: 2607.27025 by the authors.

Figure 1
Figure 1. b). A possible mapping between the pair {i, j} and an integer n is n = 1 + 3(i + 1) + (j + 1); so that, for example, {i, j} = {1,1} → n = 9 and {i, j} = {−1,−1} → n = 1. Since the final measurements occur on the detectors, we trace out the degrees of freedom of S, corresponding to taking j ′ = j in Eq. (1) (see Appendix). With the notation change, the detectors’ density matrix becomes ρ¯f = 1 N ∑An,ke i(λαn−λ ′αk) e… view at source ↗
Figure 2
Figure 2. FIG. 2. Optimized output distributions from [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Pith tools

Reviewed July 30, 2026 · model on record in the stance chip above.