{"id":"19705998-d203-464d-bf2b-5a468f53e680","arxiv_id":"2411.13060","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A two-qubit entangled state was teleported cyclically around a reused ring of ion-trap qubits, preserving entanglement after 56 hops, beyond the device's 20 qubits.","lead":"Researchers teleported an entangled two-qubit state around a loop of qubits on a 20-qubit ion-trap quantum computer, reusing the same physical qubits after each step. The state kept measurable entanglement after 56 hops, more hops than the number of qubits on the chip, a step toward measurement-based quantum computing.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Mid-circuit readout errors in dynamic-circuit feedforward are uncalibrated and unmitigated, yet the reported error bars exclude them; a real-machine dynamic-vs-post-selection comparison would quantify their impact.","rationale":"The central claim of sustained entanglement beyond 20 qubits rests on the 56-hop dynamic-circuit experiment. The correction loop is the element that makes the reuse protocol distinct from ordinary teleportation, and it is also the element whose correctness depends on mid-circuit measurement outcomes. The paper's own Figure 3 caption concedes that readout error mitigation is applied only to the final two qubits, so the raw mid-circuit outcomes are unmitigated and their error rates are not characterized. This is a genuine gap in the uncertainty budget, not a question of consensus: the numerical result could still be correct, but its error bars do not include a plausible error source that directly affects the feedforward mechanism. The mismatch between the dynamic-circuit and post-selection approaches would expose this directly because post-selection avoids the feedforward hardware altogether while sharing the same mid-circuit classification dependence. The proposed test is concrete and could settle whether the concern actually lands. The reader's weakest_assumption identifies the same issue, so agreement is 'agree'. Since the reader already issued a conditional verdict with this as one of the conditions, my read does not move the verdict; it reinforces the condition.","tokens_in":9727,"tokens_out":17376,"duration_ms":172658,"concrete_test":"On H1-1, run the 56-hop protocol in two modes: (a) dynamic circuits exactly as reported, and (b) the same circuit with all feedforward gates removed, then reclassify each shot in software by the measured mid-circuit outcome into the four byproduct classes (Eq. 4) and construct each class density matrix via QST. Compare the class-average negativity to mode (a). Since mode (b) never applies corrections, any significant difference isolates the effect of erroneous feedforward; if the two agree within the bootstrapped error bars, mid-circuit readout errors are not load-bearing. Additionally, insert a calibration block that prepares |+> and measures it in X under the same mid-circuit timing to estimate the per-measurement error rate p, and propagate (1-p)^56 to bound the worst-case negativity shift.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The byproduct correction in Eqs. (4)-(5) is contingent on the classical outcomes s_i of each mid-circuit X measurement being correct. A single readout error flips the parity used to choose the feedforward gate, so the applied correction is the wrong Pauli (or Hadamard) and the shot ends in a different locally-equivalent graph state. Averaging over shots, this creates a mixture of Bell-diagonal states; for a per-measurement error rate p, the final fidelity to the ideal state falls as (1-p)^m, and the negativity can drop well below the ideal 0.5. The paper explicitly states (Figure 3 caption) that REM is applied only to the two final qubits, so the mid-circuit outcomes are unmitigated and their error rate is not reported. The headline 0.291±0.018 at 56 hops therefore has an unquantified contribution from this feedforward error channel; the bootstrapped error bars cover only final-shot statistical noise. If the mid-circuit readout fidelity is lower than assumed, the margin above the separability threshold shrinks, and the central 'entanglement preserved beyond 20 qubits' claim is less secure.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a cyclic quantum teleportation protocol, the 'quantum hamster wheel', in which a two-qubit graph state is teleported repeatedly around a ring of qubits by measuring intermediate qubits, resetting them, and re-entangling them for further hops. The protocol is tested on the 20-qubit Quantinuum H1-1 processor and its noise-model emulator, with dynamic-circuit feedforward correction of the byproduct operators. The authors report a real-machine negativity of 0.291 ± 0.018 after 56 hops and an emulator negativity of 0.224 ± 0.009 after 100 hops, and interpret these results as demonstrating entanglement preservation beyond the number of available qubits. The mathematical framework (Eqs. 3-5) is standard stabilizer teleportation and is presented clearly.","tokens_in":9799,"tokens_out":5280,"duration_ms":52141,"significance":"If the experimental claims are fully supported, the paper would provide a concrete demonstration of a reusable, measurement-based teleportation resource that exceeds the physical qubit count of the device, which is a useful step toward measurement-based quantum computation. Strengths of the paper include the clean formulation of the byproduct operator, the use of both dynamic-circuit and post-selection implementations, the availability of source code, and the explicit acknowledgment of several experimental limitations. However, the central quantitative claims rest on experimental data whose error bars omit a potentially dominant error source: uncalibrated mid-circuit readout errors in the dynamic-circuit feedforward. The emulator results also contain a discontinuity from a change in noise-model parameters. These issues must be addressed before the headline numbers can be taken at face value.","major_comments":[{"comment":"The dynamic-circuit results are vulnerable to mid-circuit readout errors that are neither calibrated nor mitigated. The Figure 3 caption explicitly states that REM is applied only to the two qubits of the final teleported state, yet the feedforward correction in Eqs. (4)-(5) depends on all mid-circuit measurement outcomes. A single erroneous outcome flips the discriminator parity and applies the wrong byproduct operator, leaving an uncorrected Pauli error on the teleported state. The reported error bars, obtained by bootstrapping the final QST counts, account only for shot noise and exclude this feedforward error channel. The authors should report the device's mid-circuit readout fidelity and either fold it into the uncertainty estimate or provide a real-machine comparison between dynamic circuits and post-selection at the same hop counts to bound the feedforward penalty.","section":"Results / Figure 3 caption"},{"comment":"The normalization of the post-selection density matrix is unclear or incorrect as written. If the sum runs over all 2^m measurement outcome strings s, the prefactor 1/2^{m-2} gives a trace of 4 rather than 1. If the sum is intended to run only over outcomes within one discriminator class, that class contains 2^{m-2} strings and the prefactor is correct, but the text does not state this restriction. Since the post-selection negativity and fidelity values are presented as evidence in Figure 3, the definition of the averaged state must be made precise and correct.","section":"Eq. (8)"},{"comment":"The real-machine data consist of a single trial per hop count (9, 18, and 56 hops), and the error bars are obtained by bootstrapping the 1000-shot QST measurement within that single trial. This quantifies only within-run statistical fluctuations, not run-to-run reproducibility or calibration drift. The claim '0.291 ± 0.018 after 56 hops' should therefore be presented as a single-run result, and at least one repeated run at a given hop count would be needed to assess the reliability of the reported error bars.","section":"Results / Figure 3"},{"comment":"The emulator data show a clear discontinuity at 76 hops, which the authors attribute to a change in noise-model parameters due to a time gap between experiments. The 100-hop negativity of 0.224 ± 0.009 is obtained after this parameter change and lies on a different noise baseline than the earlier hop counts. Consequently, the statement that 'the actual limit of the number of hops ... is far beyond 100' is not supported by a consistent noise-model trend. The authors should either rerun the emulator with a single calibration across all hop counts or explicitly limit the extrapolation claim.","section":"Results / Figure 3 and Discussion"}],"minor_comments":[{"comment":"The abstract says 56 hops is 'three complete revolutions around the hamster wheel,' but with a ring of 19 qubits, three full revolutions would be 57 hops; the equivalence to 56 hops should be explained or corrected.","section":"Abstract and Discussion"},{"comment":"The fidelity formula contains an unusual notational artifact, '− − − − − →', and the condition 'tr(σ)=1' is used in a confusing way; this should be rewritten cleanly.","section":"Eq. (7)"},{"comment":"The caption states both that 'All results are mitigated with REM' and that 'REM is applied to only the two qubits of the final teleported state' for the dynamic-circuit approach; this apparent contradiction should be clarified in the main text.","section":"Figure 3 caption"},{"comment":"The y-axis of panel (b) is labeled 'Negativity' but the panel shows fidelity; this should be corrected to 'Fidelity'.","section":"Figure 4"},{"comment":"The text says the code is available in the GitHub repository [29], but the Data Availability Statement says data are available from the corresponding author upon reasonable request; these statements should be reconciled.","section":"Data Availability Statement"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper does something real and modestly new: it adapts graph-state teleportation to a regenerating ring, using resets to reuse qubits, and shows on Quantinuum H1-1 that a two-qubit entangled state survives 56 hops—more than the 20 physical qubits. The mathematics in Eqs. (3)-(5) is standard stabilizer teleportation and looks correct. The byproduct operator formula is explicit, and the dynamic-circuit versus post-selection comparison on the emulator is a nice touch. The reported negativities and fidelities are measured, not fitted, and the source code is on GitHub. That is genuine evidence of a useful subfield-level demonstration.\n\nThe soft spots are real but not fatal. The strongest concern, also flagged in your stress-test, is that mid-circuit readout errors are unquantified and unmitigated. REM is applied only to the final two qubits, so a wrong mid-circuit outcome directly produces the wrong byproduct operator in the dynamic-circuit case. The same errors corrupt the post-selection classification. The bootstrapped error bars cover only final-shot statistical noise, not this feedforward error channel. The paper does not report mid-circuit readout fidelity, and the magnitude of the effect could be small on H1-1, but it should be measured. A real-machine dynamic-versus-post-selection comparison would settle it.\n\nSecond, the real-machine data are single-trial per hop count. Bootstrapping within one run gives error bars on shot noise, not on run-to-run variability. That limits the strength of the 0.291 ± 0.018 headline. Third, Eq. (8) is ambiguous: it sums over all outcomes, which would give a mixture over all four variants, yet the text says the density matrix is constructed per variant. A clarifying index restriction is needed. Fourth, the emulator data have a discontinuity at 76 hops due to a noise-model change; the authors acknowledge this, but it makes the 100-hop trend less clean.\n\nNone of these issues sinks the central argument. The protocol is coherent, the data support the claim of entanglement beyond 20 qubits, and the paper is transparent about its limitations. The main missing piece is a characterization of mid-circuit readout errors and their impact on the byproduct correction. The paper deserves a serious referee, but it needs revision and ideally additional data before acceptance.\n\nFor the reading group: it is a worthwhile example of dynamic circuits in MBQC, though not a conceptual breakthrough. I would not cite it in my own work within the next year, but I would send it to review.","headline":"A solid, honestly reported demonstration of cyclic graph-state teleportation beyond the device qubit count; the main unquantified risk is the effect of mid-circuit readout errors on the feedforward corrections.","tokens_in":10484,"tokens_out":1937,"would_cite":false,"duration_ms":21290,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81P68","81P40"],"pacs":["03.67.-a","03.67.Mn"],"model":"deepseek-v4-flash","headline":"This paper shows that a two-qubit graph state can be teleported around a ring of qubits again and again, because every measured qubit is reset and re-entangled, so the teleportation distance is no longer capped by the number of physical…","keywords":["quantum teleportation","graph states","measurement-based quantum computation","dynamic circuits","qubit reuse","entanglement negativity","readout error mitigation","ion-trap quantum computer"],"falsifier":"A concrete test: inject a known readout error model into the emulator's mid-circuit measurements, for example flip each outcome with probability $p$, and measure the negativity after 56 hops; if a modest $p$ destroys the entanglement, then the demonstrated hop depth rests on unverified mid-circuit readout fidelity.","tokens_in":9390,"feed_emoji":"🌀","tokens_out":7937,"duration_ms":78985,"temperature":0.7,"pith_summary":"This paper claims that quantum teleportation can be made reusable: instead of sending a two-qubit entangled state once along a chain and losing the used qubits, the authors reset and re-entangle every measured qubit, turning a line into a regenerating ring they call a hamster wheel. With this scheme, the teleported state can hop around the ring many times, exceeding the total number of qubits in the hardware. On a 20-qubit ion-trap computer the two-qubit graph state remained entangled after 56 hops, three full revolutions, with a negativity of $0.291 \\pm 0.018$, about 58% of the maximum, and the emulator showed entanglement surviving 100 hops. The significance is that teleportation depth becomes a matter of accumulated noise and correction quality rather than device size, which speaks directly to measurement-based quantum computing, where computation proceeds by measuring an entangled resource state.","feed_headline":"Teleported state survives 56 hops on a 19-qubit wheel","feed_subtitle":"Measured qubits are reset and re-entangled, letting entanglement persist long past the hardware's qubit count.","key_machinery":"The central object is the regenerating ring graph state, a path graph whose measured vertices are reset to $|0\\rangle$, restored to $|+\\rangle$, and re-entangled with the current receiver via controlled-Z gates so the path reforms in a new cycle. The identity that carries the argument is the accumulated byproduct operator $U_m = H^m Z^{\\oplus_{\\text{odd}} s_i} X^{\\oplus_{\\text{even}} s_i}$ on the moving qubit, whose correction requires only the XOR of measurement outcomes, computed by the discriminator. This reduces the correction to a fixed-depth dynamic circuit of at most three gates, independent of hop count, and explains why dynamic correction and post-selection perform almost identically.","core_discovery":"Starting from a two-qubit graph state $|\\phi(P_2)\\rangle_{0,1}$ entangled with a line of qubits prepared in $|+\\rangle$ and linked by controlled-Z gates, the authors teleport the second qubit around the ring by measuring successive qubits in the Pauli-X basis. Each hop leaves the teleported state changed by a local unitary $H^m Z^{s_1\\oplus s_3\\oplus\\cdots} X^{s_2\\oplus s_4\\oplus\\cdots}$ whose exponents are the parities of the measurement outcomes from odd and even hops. After measuring all ring qubits except the axis and receiver, they reset the measured qubits and re-entangle them with the receiver, allowing further hops around the same ring. The local transformation is undone either by a byproduct operator applied through a dynamic circuit conditioned on the mid-circuit outcomes, or by post-selecting on the four possible outcome classes. Using quantum state tomography with readout error mitigation, they report negativity $0.291\\pm0.018$ and fidelity $0.791\\pm0.018$ after 56 hops on the real device, and negativity $0.224\\pm0.009$ after 100 hops on the emulator, showing that entanglement is preserved across more hops than the device has qubits.","pith_inferences":["If mid-circuit readout errors are the dominant uncorrected error source, as the paper's own mitigation scheme suggests because readout error mitigation is applied only to the final two qubits, then improving intermediate measurement fidelity should raise the hop ceiling more than improving two-qubit gates.","The ring-reuse idea could be generalized from a two-qubit graph state to multi-qubit encoded states or to a continuous stream of Bell pairs, effectively converting a fixed hardware register into a renewable entanglement resource for quantum repeaters.","The same cycling could implement logical gates by choosing measurement bases adaptively, turning the hamster wheel into a universal measurement-based quantum computing fabric whose depth is limited only by noise.","A direct test would be to run the scheme with randomized mid-circuit readout errors injected in the emulator and compare the negativity decay, quantifying how much of the observed decay comes from intermediate measurements rather than two-qubit gates."],"forward_implications":["Teleportation distance is no longer bounded by the number of physical qubits; any sufficiently low-noise device can in principle teleport a state arbitrarily far by cycling the ring.","Dynamic-circuit correction and post-selection give nearly the same entanglement, so the choice between them can be made on latency and shot budget rather than quality.","The slow, roughly linear decay of negativity with hops on the emulator implies that noise, not the protocol's structure, sets the hop limit; improving gate fidelity and mid-circuit readout directly extends the reachable depth.","The agreement between emulator and real device across the tested range supports using the emulator to predict performance at greater depths, such as beyond 100 hops.","The protocol is a direct resource for measurement-based quantum computation, since teleportation along a graph state by X-basis measurements is the basic operation of that model."],"supporting_citations":[{"why":"Supplies the original quantum teleportation protocol that this work extends to cyclic reuse of qubits.","marker":"[1]"},{"why":"Provides the 19-qubit teleportation protocol and dynamic-circuit discriminator that the hamster wheel adapts and reruns cyclically.","marker":"[6]"},{"why":"Gives the measurement-based quantum computation perspective that frames teleportation hops as gate operations.","marker":"[13]"},{"why":"Establishes the cluster-state measurement-based computation model that the regenerating ring realizes in miniature.","marker":"[15]"},{"why":"Supplies the device and emulator noise model used for the real-machine and simulated experiments.","marker":"[16]"},{"why":"Provides the quantum state tomography procedure used to reconstruct the teleported state's density matrix.","marker":"[23]"},{"why":"Provides the readout error mitigation method applied to the final measured counts.","marker":"[24]"},{"why":"Supplies the maximum-likelihood nearest physical density matrix algorithm used to remove unphysical tomography results.","marker":"[28]"}],"fun_headline_variants":["Quantum hamster wheel teleports qubits 56 times on a 20-qubit ring","Cyclic teleportation reuses qubits, reaching 56 hops on one ring","Entanglement survives 56 hops on a regenerating qubit wheel","Qubit reuse wheels teleportation past the 20-qubit limit","Regenerated qubits allow 56-hop teleportation on a 20-qubit ring"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Every mid-circuit measurement that decides a correction gate must return the correct classical bit; a single wrong bit applies the wrong correction and leaves an uncorrected error on the teleported state.","fun_headline_variants_meta":{"raw":{"variants":["Quantum hamster wheel teleports qubits 56 times on a 20-qubit ring","Cyclic teleportation reuses qubits, reaching 56 hops on one ring","Entanglement survives 56 hops on a regenerating qubit wheel","Qubit reuse wheels teleportation past the 20-qubit limit","Regenerated qubits allow 56-hop teleportation on a 20-qubit ring"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000865,"raw_usage":{"total_tokens":3806,"prompt_tokens":1056,"completion_tokens":2750,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":672,"completion_tokens_details":{"reasoning_tokens":2644}},"tokens_in":672,"tokens_out":2750,"duration_ms":20121,"temperature":1.0,"reasoning_tokens":2644,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T16:53:20.907811+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete test: inject a known readout error model into the emulator's mid-circuit measurements, for example flip each outcome with probability $p$, and measure the negativity after 56 hops; if a modest $p$ destroys the entanglement, then the demonstrated hop depth rests on unverified mid-circuit readout fidelity.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the 19-qubit teleportation protocol and dynamic-circuit discriminator that the hamster wheel adapts and reruns cyclically."},{"cited_title":"Jozsa, An introduction to measurement based quan- tum computation, NATO Science Series, III: Computer and Systems Sciences","cited_arxiv_id":null,"evidence_quote":"Gives the measurement-based quantum computation perspective that frames teleportation hops as gate operations."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the device and emulator noise model used for the real-machine and simulated experiments."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the quantum state tomography procedure used to reconstruct the teleported state's density matrix."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the maximum-likelihood nearest physical density matrix algorithm used to remove unphysical tomography results."}],"review_version":1}