{"id":"b2025cc9-f761-4016-8b3d-454beb9ff89d","arxiv_id":"2607.27168","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"NEGF scattering for 2- and 6-site quantum dragons is mapped to 3–4 qubit HHL/VQLS circuits; ideal runs recover the solution, while NISQ runs recover T(E)=1 mainly by ansatz constraint or noise.","lead":"Researchers ran the first quantum-computer versions of NEGF electron-transport calculations on tiny “quantum dragon” nanodevices that theoretically transmit every electron. Ideal simulations work; noisy hardware and variational runs recover perfect transmission largely by construction or noise artifacts, so the result is a compact proof-of-concept, not yet a practical transport solver.","discovery_kind":"new_application","skeptic_critique":{"model":"grok-4.5","headline":"VQLS recovers T(E)=1 by ansatz constraint, not by unconstrained solution of the NEGF linear system.","rationale":"The reader correctly isolated the load-bearing soft spot: VQLS builds T=1 into the ansatz, so recovering perfect transmission does not demonstrate an unconstrained NEGF solve. Ideal HHL overlaps and the compact similarity-reduced mapping remain genuine contributions and justify a CONDITIONAL rather than REJECT verdict; the paper is a valid first quantum-NEGF benchmark on toy dragons provided the circularity is foregrounded. No deeper internal inconsistency (e.g., algebraic error in the block-diagonalization) was found. The concrete unconstrained-VQLS rerun would settle the issue cleanly without requiring new hardware.","tokens_in":30174,"tokens_out":610,"duration_ms":24185,"concrete_test":"Re-run the 2-site VQLS pipeline of Sec. V.B on the same energy grid and shot budget, but drop the |r|²=|s|² magnitude constraint so that only CL(α) is minimized. Report both unconstrained T(E)=|s|² and |⟨x_VQLS|x⟩|². If T deviates from 1 by more than shot noise or the overlap collapses relative to Fig. 12, the claim that VQLS solves the NEGF LSE (rather than projects onto the dragon manifold) is not supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper’s central demonstration—that HHL/VQLS “compute the transmission coefficient T(E)” of quantum dragons—leans on recovering T(E)=1. For VQLS this is not an inference from solving A|x⟩=|b⟩. Section IV (after Eq. 50) states that the authors “supplement the local cost function CL(α) by a physical constraint for perfect transmission |r|²=|s|² which requires the αr and the αs component of the trial state |α⟩ to have the same magnitude,” and later (V.B) that “by construction, T(E)=1 is recovered across the entire conducting bandwidth regardless of the optimizer’s wavefunction fidelity.” Thus the headline observable is enforced on the variational manifold rather than predicted by an unconstrained solve. Noisy HHL likewise prints T=1 because decoherence flattens the block-register probabilities to 1/8 (discussion of Fig. 10), so the ratio |xs|²/|xr|² is trivially unity. The only non-circular evidence is wavefunction overlap with the classical solution, which is strong only for ideal HHL and mid-band VQLS and degrades where κ is large (Figs. 12–14). Choosing devices already known to satisfy the dragon eigenvalue condition (Eqs. 17, 26) further softens the claim that the algorithms are discovering transport physics.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The manuscript recasts the NEGF scattering problem for single-band tight-binding nanodevices as a linear system A|x⟩=|b⟩ whose solution encodes reflection and transmission amplitudes, then applies HHL and VQLS to 2-site and 6-site quantum-dragon devices that are known to satisfy T(E)=1 across the lead band. A similarity transformation block-diagonalizes the system into an equivalent linear chain plus a null subspace, reducing the Pauli decomposition and mapping the problems onto 3–4 physical qubits. Ideal HHL statevector overlaps reach ~99% and ideal shot simulations recover T≈1; noisy HHL collapses; VQLS (with a |r|²=|s|² ansatz constraint) yields mid-band overlaps ≳90% on ibm_torino for the 2-site device and weaker ideal-simulation overlaps for the 6-site device.","tokens_in":30494,"tokens_out":1522,"duration_ms":32803,"significance":"If the claims hold under a clarified reading, this is a useful first demonstration that NEGF transport LSEs can be mapped to compact HHL/VQLS circuits, with an explicit similarity reduction that lowers Pauli weight. Ideal HHL fidelity and a limited hardware VQLS run on ibm_torino are concrete benchmarks. The work does not claim a complexity advantage for these tiny systems; its value is methodological (NEGF→LSE→quantum linear algebra) and as a pathway toward larger nanodevice Hamiltonians on future FT hardware. Strengths include the careful classical reformulation (Secs. II–III), the block-diagonalization that reduces L, and transparent reporting of noisy-HHL failure modes.","major_comments":[{"comment":"Sec. IV (after Eq. 50) and Sec. V.B: VQLS supplements CL(α) with the constraint |r|²=|s|² on the trial state, so T(E)=1 is recovered “by construction… regardless of the optimizer’s wavefunction fidelity.” The abstract and introduction frame the result as computing T(E) of quantum dragons. That framing overstates what VQLS demonstrates. The load-bearing evidence is wavefunction overlap with the classical solution (Figs. 12–14), not T(E). The manuscript should restate the VQLS claim around overlap/fidelity of |x⟩, treat T=1 as an imposed physical prior, and report unconstrained VQLS (or an ablation) so readers can separate ansatz bias from solver performance.","section":"Section IV, Section V.B"},{"comment":"Sec. V.A and discussion of Fig. 10: noisy HHL yields ΔHHL≈0.375 because decoherence flattens block-register probabilities to 1/8, which forces the ratio |xs|²/|xr|²=1 and thus T=1 trivially. Presenting noisy T(E)≈1 alongside ideal T(E) without equal emphasis that the observable is uninformative under that noise model weakens the “feasibility on physical processors” claim for HHL. Either omit noisy T as a success metric or state explicitly that only overlap (and structured probability vectors) are diagnostic.","section":"Section V.A, Fig. 10"},{"comment":"Secs. II.B–II.C, Eqs. (17), (26), (29): devices are constructed to satisfy the dragon eigenvalue conditions from prior work, so T(E)=1 is an input property, not a prediction. That is legitimate for benchmarking solvers, but the paper should say clearly that the algorithms are validated on instances with known closed-form structure (equivalent linear chain + null modes), not that they discover ballistic transport in disordered systems. Without that caveat, the “training quantum dragons” narrative reads stronger than the experiments support.","section":"Sections II.B–II.C, Eqs. (17), (26)"},{"comment":"Figs. 12–14 and Sec. V.D: VQLS overlap degrades sharply where the condition number κ of the reduced block is large (band edges), and the 6-site device (L=14, larger Hilbert space) shows substantially worse ideal overlap than the 2-site case. The comparison that “HHL is unaffected by κ while VQLS is not” is only supported in ideal simulation; noisy HHL is unusable. A revised discussion should quantify the κ window where VQLS is reliable, state optimizer budgets (Powell evaluations, shots) as limiting factors, and avoid implying a clean NISQ-vs-FT division beyond what the data show.","section":"Section V.B–V.D, Figs. 12–14"}],"minor_comments":[{"comment":"Abstract: “3 and 4 total physical qubits” is easy to misread; HHL uses additional clock and flag qubits (nc≈6–7 plus ancilla). Clarify system vs total qubits.","section":"Abstract"},{"comment":"Eq. (36) and surrounding text: normalization constant c and the relation T=|xs|²/|xr|² under current conservation should be cross-checked against the block-encoded |x̃⟩ convention in Eq. (39) so readout formulas are unambiguous.","section":"Section II.D, Eq. (36)"},{"comment":"Table II: report the achieved CL values and whether termination was by budget or convergence; “Iterations” alone is hard to interpret.","section":"Table II"},{"comment":"Typos/notation: “wavefucntion” (II.A); “V ARIA TIONAL” spacing in the Sec. IV heading; “ibmq torino” vs “ibm_torino” inconsistency; “QVLS” vs “VQLS” in Fig. 12 caption.","section":"Throughout"},{"comment":"Related work: Yang et al. [53] is noted; a short explicit contrast (dragon vs general devices; constrained vs unconstrained cost) would help priority and scope.","section":"Section V.D"},{"comment":"Appendix B: t-tuning via expected 1/√3 probabilities is useful; state whether this protocol was used only in simulation or is proposed for hardware (cost).","section":"Appendix B"}],"recommendation":"major_revision","confidential_remarks":"The skeptic’s circularity point on VQLS is correct and load-bearing for how the abstract is written; it is fixable by reframing without new theory. I would not reject: the NEGF→LSE mapping and ideal-HHL benchmarks are real contributions for quant-ph / nanoelectronics. Fit is appropriate for a quantum-algorithms or computational-physics venue; novelty is “first NEGF on QC for dragons,” not a complexity result. No integrity concerns."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The real news is narrow but real: they turn NEGF scattering for quantum dragons into a small LSE, block-encode it, shrink the Pauli weight with the dragon similarity/permutation, and run HHL and VQLS on 3–4 physical qubits. That is the first explicit quantum-circuit NEGF pipeline I have seen, not a rehash of dragons or of HHL/VQLS alone.\n\nWhat they do well is the classical setup. The scattering LSE, non-Hermitian block encoding, and reduction to a linear chain plus null subspace match known dragon theory and are written carefully enough to check. Ideal HHL statevector overlaps sit near 99% and shot simulations track T≈1 with sensible success probabilities. Mid-band VQLS overlaps on the 2-site device are decent in ideal/noisy sims and on a few ibm_torino runs. The Pauli-count cut from dropping the null block is a practical detail worth keeping.\n\nSoft spots, in proportion. VQLS supplements the local cost with |r|²=|s|² on the trial state, so T=1 is on the manifold by construction; the paper even says fidelity can tank while T stays 1. Noisy HHL also prints T=1 because the block probabilities flatten to 1/8. The non-circular evidence is wavefunction overlap, which is strong only for ideal HHL and mid-band VQLS and falls where κ is large. Devices are pre-chosen dragons, so the algorithms are not discovering transport physics—they are recovering a known solution on 2- and 6-site single-band models. Hardware is a handful of 2-site VQLS points; 6-site is ideal sim only. No code dump.\n\nThis is for people who care about quantum linear algebra for nanoelectronics or want a concrete NISQ/FT benchmark path (VQLS now, HHL later). It is not a materials result and not a noisy-device transport solver. Math and citations look solid; circularity is real but localized to how T is sold under VQLS.\n\nI would send it to peer review. Ask them to foreground that T=1 is constrained under VQLS, lead with overlaps, and keep the scale honest. Worth engaging as a methods paper, not as a physics discovery.","headline":"Solid first quantum-NEGF circuit demo on toy dragons; T=1 under VQLS is mostly enforced by the ansatz, so judge it on overlaps and the mapping, not the headline transmission.","tokens_in":31174,"tokens_out":587,"would_cite":true,"duration_ms":18167,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"NEGF electron transport for quantum-dragon nanodevices can be solved on quantum computers with only three to four physical qubits.","keywords":["quantum computing","HHL","variational quantum linear solver","NEGF","electron transport","quantum dragon","nanodevices","tight-binding model"],"falsifier":"Run unconstrained VQLS (no |r|²=|s|² forced on the ansatz) on the same 2-site dragon matrices and check whether the measured |s|² still equals 1 across the band and whether the full state overlaps the classical dragon solution above the noise floor.","tokens_in":30983,"feed_emoji":"⚛️","tokens_out":965,"duration_ms":22688,"temperature":0.7,"pith_summary":"This paper shows that the standard NEGF method for nanoscale electron transport can be rewritten as a linear system whose solution carries the transmission and reflection amplitudes, making it runnable on quantum linear-algebra algorithms. The authors apply HHL and the variational quantum linear solver to quantum-dragon devices—nanodevices that transmit every electron in the conducting band even when the interior is disordered—and recover the perfect-transmission signature T(E)=1. After a similarity transformation that block-diagonalizes the system, the 2-site and 6-site dragons fit on compact circuits of three and four physical qubits. Ideal simulations, noise-aware simulations, and runs on an IBM processor are used to argue that the approach is already feasible on near-term hardware for VQLS and is the natural path for fault-tolerant HHL as devices improve. A sympathetic reader cares because NEGF is the workhorse of nanoelectronics design, and a quantum route could eventually reach disordered or correlated devices that classical solvers struggle with.","feed_headline":"Quantum chips solve dragon nanodevice transport on 3–4 qubits","feed_subtitle":"NEGF scattering is recast as a linear system; HHL and VQLS recover perfect transmission T(E)=1","key_machinery":"A similarity transformation that block-diagonalizes the NEGF linear system, decoupling the perfectly transmitting dragon mode from null modes and sharply reducing the Pauli decomposition of the block-encoded matrix so the problem fits compact HHL and VQLS circuits.","core_discovery":"The first quantum-computer implementation of NEGF is obtained by recasting the scattering problem as a linear system solvable by HHL and VQLS; for quantum-dragon nanodevices in the single-band tight-binding model, those algorithms recover the perfect-transmission solution T(E)=1 on circuits of only three to four physical qubits, as shown in ideal and noise-aware simulations and on physical IBM hardware.","pith_inferences":["Because the dragon condition is an eigenvalue constraint that forces a zero mode, any quantum linear solver that preserves that kernel will automatically report T=1; the harder test is devices that are not dragons, where transmission is energy-dependent and cannot be hard-wired into the ansatz.","The extreme compactness (3–4 qubits) is bought by the dragon similarity transform; without it, general disordered NEGF matrices would need far more Pauli terms and deeper circuits, so the headline qubit counts do not automatically transfer to arbitrary nanodevices.","If unconstrained VQLS fails near band edges while constrained VQLS succeeds, the method is better read as a constrained variational search for dragon-compatible states than as a general NEGF inverter."],"forward_implications":["VQLS is positioned as the practical near-term route for NEGF transport on NISQ processors, while HHL is the intended fault-tolerant successor once logical error rates fall.","The same block-diagonal dragon reduction can be reused for larger multi-slice devices that share graphene or carbon-nanotube connectivity.","Physics-aware ansätze that encode current conservation can keep the physical observable T(E)=1 even when wavefunction fidelity degrades near band edges.","Extensions to DNA molecular wires and full tight-binding nanotube junctions become the natural next targets once the qubit count and optimizer scale."],"fun_headline_variants":["NEGF on 3–4 qubits recovers T(E)=1 for quantum dragons","HHL and VQLS solve dragon nanodevice scattering as linear systems","First quantum NEGF maps dragon transport to 3–4 qubit circuits","Block-diagonal NEGF yields perfect transmission on IBM hardware","Quantum dragons keep T(E)=1 under disorder via HHL/VQLS"],"cache_read_input_tokens":28288,"weakest_assumption_plain":"That recovering T(E)=1 with the variational solver shows the linear system was solved, even though perfect transmission is built into the trial state by a hard constraint rather than left free for the optimizer to discover.","fun_headline_variants_meta":{"raw":{"variants":["NEGF on 3–4 qubits recovers T(E)=1 for quantum dragons","HHL and VQLS solve dragon nanodevice scattering as linear systems","First quantum NEGF maps dragon transport to 3–4 qubit circuits","Block-diagonal NEGF yields perfect transmission on IBM hardware","Quantum dragons keep T(E)=1 under disorder via HHL/VQLS"]},"model":"grok-4.5","effort":"low","cost_usd":0.002059,"raw_usage":{"total_tokens":899,"prompt_tokens":725,"num_sources_used":0,"completion_tokens":83,"cost_in_usd_ticks":20588000,"prompt_tokens_details":{"text_tokens":725,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":91,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":725,"tokens_out":83,"duration_ms":3239,"temperature":1.0,"reasoning_tokens":91,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-30T10:58:23.844847+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Run unconstrained VQLS (no |r|²=|s|² forced on the ansatz) on the same 2-site dragon matrices and check whether the measured |s|² still equals 1 across the band and whether the full state overlaps the classical dragon solution above the noise floor.","supporting_citations":[],"review_version":2}