{"id":"149b4cb6-3ebe-4027-9486-0a738bc3ae81","arxiv_id":"2412.12959","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":7,"one_line_summary":"A room-temperature three-nuclear-spin NV register is characterized by randomized benchmarking, and a noise model fit to those errors produces a simulated quantum volume of 8.","lead":"Researchers measured gate errors on the three nuclear spins around a nitrogen-vacancy center in diamond at room temperature. From those errors they built a noise model and simulated a quantum volume of 8, the largest possible for a three-qubit system.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The QV=8 headline is a simulated result: the RB-calibrated depolarizing model is never validated on actual QV circuits, and the 15 kHz data-discard rule removes an acknowledged drift error; a direct m=3 heavy-output run is the missing check.","rationale":"The reader's CONDITIONAL verdict already identifies the correct load-bearing weakness: the QV=8 result rests on a calibrated noise model whose fidelity for longer, differently structured QV circuits is not established, and the 15 kHz data-discard rule removes an acknowledged error source. My stress-test pass finds no additional internal inconsistency in the RB formalism or in the QV simulation procedure; the measured EPGs are standard, and the simulation steps are clearly described. The concern is external validity of the model, not mathematical soundness of the RB extraction. I agree with the reader that a direct heavy-output QV run is the decisive missing check. The fact that QV=8 is the maximum possible for three qubits further reduces the significance of the value even though it does not invalidate the modeling claim. The paper is honest in its title about modeling, but the body uses 'achieving' language that blurs simulation and measurement, so the conditional verdict is appropriate. No change to the reader's verdict is needed.","tokens_in":10837,"tokens_out":4860,"duration_ms":50006,"concrete_test":"Run the actual m=3 quantum-volume experiment on the NV register: generate a sample of 100 random width-3 depth-3 QV circuits, execute each on the hardware, and compute the measured heavy-output probability against the ideal-circuit median, applying the same 2-sigma criterion. Repeat the analysis twice: once with the 15 kHz ODMR-shift data-discard rule in place and once without it. If the heavy-output pass only occurs after discarding the shifted runs, the simulated QV=8 overstates what the register achieves as a directly measured quantum volume.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim in Section III.C is that the register achieves quantum volume 8, but this is a prediction from a noise model fitted to randomized benchmarking data, not a measured heavy-output result. The load-bearing assumption is that the RB-calibrated per-gate depolarizing error rates transfer to the differently structured, all-three-qubits-active QV circuits. Two concrete threats undermine this assumption. First, each logical two-qubit gate is implemented by three electron-mediated CNOTs through the electron ancilla (Section III.A), yet the model assigns a single depolarizing error to the composite gate. The two-qubit RB used for calibration only exercises the two nuclear qubits in the benchmarked pair; any state-dependent crosstalk or heating that appears when all three nuclear spins are simultaneously in superposition is invisible to the fit. Second, Section III.C explicitly states that runs with ODMR frequency shifts above 15 kHz are discarded, and the text identifies heating-induced frequency shifts as a real error source for 13C2. Fitting EPGs on this post-selected subset makes the simulated QV optimistic for the raw hardware. The consistency check that the noise model 'reproduces the measurement results' validates the model only on the same RB data used to construct it, so it does not test the extrapolation to QV circuits. The claim is therefore an upper-bound-style modeled estimate. If a direct QV experiment without the favorable post-selection yields heavy-output probability below the 2/3 threshold, the central claim would be substantially overstated.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports randomized benchmarking of a room-temperature NV-center register with three nuclear-spin qubits (14N, 13C1, 13C2) connected through an electron-spin ancilla. Single- and two-qubit error-per-gate (EPG) values are extracted from exponential decay fits, an error map is constructed, and a depolarizing noise model is built in Qiskit. The authors then simulate quantum-volume (QV) circuits for widths and depths up to m = 3, obtaining a heavy-output probability above the 2/3 threshold on all qubit connections and concluding that the register has a quantum volume of 8. The entire QV result is obtained from simulation with a noise model calibrated on the RB data, not from a direct heavy-output experiment on the hardware.","tokens_in":11140,"tokens_out":2595,"duration_ms":26745,"significance":"If the reported QV = 8 is taken as a validated experimental result, it would be a useful benchmark for room-temperature NV platforms and a clean demonstration of using RB-derived EPGs to predict a composite performance metric. The manuscript has concrete strengths: standard RB protocols with explicit EPG extraction, an all-to-all connectivity map, a reproducible-looking noise-model construction, and a forward simulation of QV circuits. The forward prediction is not definitionally circular, since the QV circuits are not used to fit the model. However, the paper's central claim is substantially weaker than its wording suggests: the QV value is a model prediction with no direct experimental validation, and the model is validated only on the RB data used to construct it. That gap is load-bearing because QV is, by definition, a measured circuit-level property.","major_comments":[{"comment":"The central claim 'The quantum volume circuit for m = 3 was successful on all possible qubit connections, achieving a quantum volume of 8' is based on simulation, not on a direct heavy-output experiment on the register. The text in III.C states 'Our simulations for quantum volume circuits are performed with the noisy model' and Fig. 4(c) labels the heavy-output probability as 'simulated'. Quantum volume is defined by a threshold on measured heavy-output probabilities, so the manuscript should state clearly that QV = 8 is a modeled estimate. A direct m = 3 heavy-output experiment, even with modest statistics, would make the claim experimentally grounded and should be reported if available.","section":"Section III.C, Fig. 4(c)"},{"comment":"The consistency check that the noise model 'reproduces the measurement results of the nuclear spins' is performed on the same RB datasets from which the EPG parameters were extracted. This is an in-sample fit and does not test whether the depolarizing model transfers to longer, differently structured QV circuits that involve all three nuclear qubits simultaneously. An out-of-sample check, such as predicting RB decay at sequence lengths beyond the fitted range, or predicting a separate interleaved RB or three-qubit RB measurement, would materially support the extrapolation to QV circuits.","section":"Section III.B, noise model validation"},{"comment":"Each logical two-qubit gate between two nuclear spins is implemented by three electron-mediated CNOT gates, but the noise model assigns a single depolarizing error rate to the composite logical gate. The two-qubit RB calibration is performed on the pair under test only, so state-dependent crosstalk, leakage, or heating that appears only when all three nuclear registers are simultaneously active is invisible to the calibration. The authors should either provide evidence that the composite-gate error is independent of the state of the third nuclear spin or use a three-qubit benchmarking protocol to validate the effective two-qubit error channel.","section":"Section III.A, composite two-qubit gate model"},{"comment":"The text states that 'Measurements exhibiting a shift in the ODMR frequency of 15 kHz will be discarded' and, earlier, identifies heating-induced frequency shifts as a real error source for 13C2. Fitting the RB EPGs on this post-selected subset means the noise model describes the hardware only under a favorable operating condition. The simulated QV = 8 therefore corresponds to a post-selected (drift-free) register, not to the raw hardware as operated. The authors should quantify the bias introduced by this discard rule, for example by reporting the fraction of discarded runs and the EPG values without the cutoff, or by running a direct QV experiment without this post-selection.","section":"Section III.C and Section IV, 15 kHz ODMR shift discard rule"}],"minor_comments":[{"comment":"There are small typographical errors: 'will be refereed to as' should read 'will be referred to as', and the phrase 'will be called c 13C2' contains a stray 'c' before '13C2'.","section":"Section II"},{"comment":"Reference [26] ('Quantum volume. Nature, 2017') is incomplete and does not give authors, volume, or page numbers; please provide a complete citation or use the arXiv/preprint version of the quantum-volume proposal.","section":"Reference [26]"},{"comment":"In Eq. (4), the parameters f1 and f2 are described as 'Azz/2' but no units are given; please specify whether they are frequencies in Hz and how they relate to the hyperfine couplings in Table I.","section":"Equation (4)"},{"comment":"The T2* entry for the electron spin is given as 12.85 µs without an uncertainty, while other entries have uncertainties; adding an error bar and a brief note on how T2* was extracted would improve precision.","section":"Table I"},{"comment":"The phrase 'record quantum volume metric for ambient condition operation to date' is not supported by a comparison with other room-temperature platforms; either add a comparative table or soften the claim.","section":"Section III.C"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a solid benchmarking report, but the headline QV = 8 claim needs to be repositioned as a model prediction unless direct experimental QV data are added. The in-sample validation and the 15 kHz post-selection rule are the two main correctness risks. The paper could be suitable for the journal after a major revision that either adds an experimental QV run or clearly and consistently rewrites the claim as 'predicted quantum volume'. I do not see citation or novelty-disclosure problems beyond the incomplete reference [26]."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: the paper is a competent RB characterization of a room-temperature three-nuclear-qubit NV register, and the per-gate error rates are genuinely useful new data. The quantum-volume claim, however, is a simulation from a noise model fitted to those RB results, not a direct heavy-output measurement. That distinction matters, and the paper blurs it.\n\nWhat is new: measured single-qubit EPGs for 14N, 13C1, 13C2 and two-qubit EPGs for all three pairs, with the 13C2 90 kHz coupling and heating drift. That is a clean experimental dataset, and the error map is a nice compact summary. The QV estimate for a room-temp NV register is new in the sense that no one has published one, but QV=8 is the ceiling for three qubits, so as a metric it doesn't add much beyond saying the register is coherent enough to run depth-3 circuits in simulation.\n\nWhat the paper does well: RB is standard and the fits look reasonable; the error model built from EPGs plus readout/init errors reproduces the RB decay curves, which is an honest but weak check since it is the same data used to fit. The 15 kHz ODMR drift discard rule is explicitly disclosed, and the text acknowledges heating as a real error source. That is better than hiding it.\n\nSoft spots, in proportion: the main claim is a modeled estimate. The noise model assigns a single depolarizing error to a logical two-qubit gate that actually comprises three electron-mediated CNOTs, and the RB calibration only ever exercises the two nuclei in the benchmarked pair. Any crosstalk or heating that appears when all three nuclei are active in QV circuits is invisible to the fit. Discarding drift-shifted runs makes the model optimistic for the raw hardware. The 'reproduces the measurement results' statement validates the model on the RB curve it was built from, so it does not test the QV extrapolation. The 'record quantum volume metric for ambient condition operation' claim is not supported because no comparison to other ambient systems is given. None of this is fatal; the paper does not pretend to have run QV on hardware, but the abstract and closing line invite that reading.\n\nWho it is for: people working on NV registers or small-platform benchmarking who want realistic EPG numbers and a template for mapping RB to a Qiskit noise model. It deserves a serious referee: the experiment is real, the modeling is transparent enough to check, and the missing direct QV run is a fixable gap, not a conceptual error. I would ask for a direct m=3 heavy-output measurement (or a clear statement that QV is simulated only) before publication.","headline":"The measured EPGs are a real contribution; the QV=8 headline is a model prediction, not a measurement, and should be presented as such.","tokens_in":11806,"tokens_out":1700,"would_cite":true,"duration_ms":15478,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Room-temperature NV register hits quantum volume 8.","keywords":["quantum volume","randomized benchmarking","NV center","diamond spin register","noise model","heavy output probability","all-to-all connectivity","room-temperature quantum computing"],"falsifier":"Perform the quantum-volume heavy-output experiment directly on the same register, without the noise-model simulation, and compare the measured probability at $m=3$ against the $2/3$ threshold; if a direct measurement fails, or only passes with the 15 kHz ODMR-shift discard applied to a substantial fraction of runs, the modeled $V_Q = 8$ would be an optimistic prediction rather than a hardware-level result.","tokens_in":10561,"feed_emoji":"💎","tokens_out":10209,"duration_ms":81436,"temperature":0.7,"pith_summary":"The paper establishes that a three-nuclear-qubit register based on a nitrogen-vacancy (NV) center in diamond, operated at room temperature, can achieve a quantum volume of 8, the maximum possible for three addressable qubits. The figure comes from a noise model calibrated with randomized benchmarking, not from running the quantum-volume circuits on the hardware directly. If the model is faithful, the register passes the heavy-output test at width and depth 3 on every qubit pair, putting a room-temperature solid-state platform on par with other small-scale quantum processors. The result matters because quantum volume is a device-agnostic benchmark, and a reliable room-temperature register would make quantum computation more accessible outside cryogenic environments.","feed_headline":"Room-temp NV register hits quantum volume 8","feed_subtitle":"Three diamond nuclear spins pass the heavy-output threshold at width and depth 3.","key_machinery":"The carrying object is the depolarizing noise model assembled from randomized benchmarking. For every basis gate, the exponential decay $F = A_0 \\cdot \\alpha^{N} + B_0$ is fitted to the measured survival probability, and the per-gate error is extracted via $EPG(n,\\alpha) = \\frac{2^n-1}{2^n}(1-\\alpha)$. Single-qubit and two-qubit errors are combined with readout and initialization error estimates, treating each logical two-qubit CNOT between nuclear spins as one depolarizing channel even though its physical implementation uses three electron-mediated CNOTs. This model is then inserted into quantum-volume circuits built from layers of random SU(4) unitaries on random qubit pairs, with success judged by a heavy-output probability threshold of $2/3$.","core_discovery":"The paper's central claim is that a room-temperature NV-center register with three strongly coupled nuclear spins ($^{14}$N, $^{13}$C$_1$, $^{13}$C$_2$) and the electron spin as an ancilla reaches a quantum volume of $V_Q = 2^3 = 8$. This is the outcome of simulating quantum-volume circuits on a noise model calibrated by randomized benchmarking: the circuits of width and depth $m=3$ succeed on all three nuclear-pair connections, keeping the simulated heavy-output probability above the $2/3$ threshold with its full $2\\sigma$ spread. The register therefore passes the maximum quantum-volume test possible for a three-qubit system, and the paper takes this as evidence that room-temperature NV centers are a viable platform for quantum computation, with the electron spin's $T_1$ and $T_2$ times acting as the dominant error sources.","pith_inferences":["A direct heavy-output experiment on this hardware could disagree with the modeled $V_Q=8$ if the depolarizing model misses non-Markovian or context-dependent errors, and the paper provides no direct check of that transfer.","The 15 kHz ODMR frequency-shift discard rule is a form of post-selection; reporting the fraction of runs discarded would quantify how much of the quantum-volume result depends on that cut.","The same recipe — randomized benchmarking into a depolarizing model, then quantum-volume simulation — could be used to predict quantum volume for other central-spin or solid-state registers; if it generalizes, it becomes a cheap design tool for comparing architectures.","For three qubits, quantum volume saturates at 8, so a more informative cross-platform comparison would include the physical overhead behind that number, such as total sequence duration and the number of microwave and radio-frequency pulses."],"forward_implications":["At quantum volume 8, the register reaches the maximum depth-3 heavy-output success allowed by its three-qubit size; larger values are impossible without adding qubits or improving the electron spin's coherence.","The calibrated noise model reproduces both the benchmarking decays and the quantum-volume simulations, so it can serve as a predictive tool for other circuit families on this register without additional experimental runs.","All-to-all nuclear-spin connectivity means any qubit pair can be entangled through the electron ancilla in a fixed number of steps, avoiding the nearest-neighbor routing overhead common on chip-based devices.","Because the electron's $T_1$ and $T_2$ are the stated limiting factors, operating the same register at cryogenic temperatures should push the quantum volume beyond 8.","The weakly coupled carbon bath cannot raise the quantum volume even with perfect additional qubits, since the electron relaxation still bounds every two-qubit gate; these spins remain candidates for quantum memory instead."],"supporting_citations":[{"why":"Provides the randomized benchmarking method for single-qubit gates that the paper's error extraction builds on.","marker":"[12]"},{"why":"Defines the scalable and robust randomized benchmarking protocol used to extract per-gate error rates.","marker":"[13]"},{"why":"Supplies the central-spin noise-model construction the paper adapts to its register.","marker":"[19]"},{"why":"Provides the simulation toolchain in which the noisy quantum-volume circuits are run.","marker":"[20]"},{"why":"Defines the quantum volume metric as the largest square circuit passing the heavy-output test.","marker":"[26]"},{"why":"Defines heavy output generation and the 2/3 success threshold for quantum volume circuits.","marker":"[28]"},{"why":"Supplies the randomized model circuits and the validation method for near-term quantum devices.","marker":"[29]"}],"fun_headline_variants":["Room-temp NV qubits reach quantum volume 8","Quantum volume 8 achieved by room-temp NV register","NV center register passes quantum volume 8 benchmark","Randomized benchmarking yields NV quantum volume 8"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result depends on the assumption that the error rates measured in randomized benchmarking transfer unchanged to the longer, differently structured quantum-volume circuits, and that discarding runs with an ODMR frequency shift above 15 kHz does not remove a failure that would appear in a direct test.","fun_headline_variants_meta":{"raw":{"variants":["Room-temp NV qubits reach quantum volume 8","Quantum volume 8 achieved by room-temp NV register","NV center register passes quantum volume 8 benchmark","Randomized benchmarking yields NV quantum volume 8"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001031,"raw_usage":{"total_tokens":4291,"prompt_tokens":842,"completion_tokens":3449,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":458,"completion_tokens_details":{"reasoning_tokens":3386}},"tokens_in":458,"tokens_out":3449,"duration_ms":22840,"temperature":1.0,"reasoning_tokens":3386,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T13:33:37.115942+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Perform the quantum-volume heavy-output experiment directly on the same register, without the noise-model simulation, and compare the measured probability at $m=3$ against the $2/3$ threshold; if a direct measurement fails, or only passes with the 15 kHz ODMR-shift discard applied to a substantial fraction of runs, the modeled $V_Q = 8$ would be an optimistic prediction rather than a hardware-level result.","supporting_citations":[{"cited_title":"Knill, D","cited_arxiv_id":null,"evidence_quote":"Provides the randomized benchmarking method for single-qubit gates that the paper's error extraction builds on."},{"cited_title":"Benchmark- ing quantum error-correcting codes on quasi-linear and central-spin processors","cited_arxiv_id":null,"evidence_quote":"Supplies the central-spin noise-model construction the paper adapts to its register."},{"cited_title":"Wood, Jake Lishman, Julien Gacon, Si- mon Martiel, Paul D","cited_arxiv_id":null,"evidence_quote":"Provides the simulation toolchain in which the noisy quantum-volume circuits are run."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the quantum volume metric as the largest square circuit passing the heavy-output test."},{"cited_title":"Complexity-theoretic foundations of quantum supremacy experiments, 2016","cited_arxiv_id":null,"evidence_quote":"Defines heavy output generation and the 2/3 success threshold for quantum volume circuits."},{"cited_title":"Validating quantum computers using randomized model circuits","cited_arxiv_id":null,"evidence_quote":"Supplies the randomized model circuits and the validation method for near-term quantum devices."}],"review_version":1}