{"id":"45178cd1-5ff1-42a7-9187-92ebccd4978c","arxiv_id":"2607.17302","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Encircling a parameter loop clockwise versus counterclockwise in a dissipative two-qubit photonic system prepares different Bell states, with extension to three-qubit GHZ states.","lead":"This paper shows that photons undergoing a simulated dissipative quantum evolution can be steered into different entangled states depending on whether the system's parameters are swept clockwise or counterclockwise in a loop. The result offers a new, noise-robust way to prepare useful quantum states for quantum information tasks.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Steady-state overlap with |10⟩ is assumed, not verified; at Δδ/g=4 the dark state is not an eigenstate of H0, so the claimed mechanism's central assumption may be the fidelity bottleneck.","rationale":"The reader's weakest_assumption correctly identifies the adiabatic/steady-state overlap as the most load-bearing point. I analyzed the Lindbladian in Eqs. (1)-(2) and found that |10⟩ is not a true dark state of the full Liouvillian because H0 mixes it with |01⟩; the steady state is only approximately |10⟩ for large |δ|/g. The paper asserts 'sufficiently large Δδ' without quantifying it, and the chosen parameters (δ/g = 4) are not obviously in the asymptotic regime. This is a concrete, checkable assumption on which the chiral mechanism's explanation rests. A direct numerical check of the steady-state overlap would settle whether the mechanism works as described. If the overlap is high, the concern is resolved; if not, the authors should report the actual bound and adjust their explanation. The reader's CONDITIONAL verdict remains appropriate because this check is missing; the verdict should not be overturned to ACCEPT without it, nor REJECT because the experimental fidelities suggest the protocol is at least partially working. The secondary issue of only 10 noise realizations is noted but is less central to the physical mechanism.","tokens_in":11649,"tokens_out":26091,"duration_ms":251418,"concrete_test":"Compute the zero-eigenvalue eigenvector ρ_ss of L in Eq. (2) at the loop corners (γ=0.18, δ=±0.04) for ξ=1, g=0.01, ϵ=1.2, and evaluate F10 = ⟨10|ρ_ss|10⟩ as well as the fidelity with the H0 eigenstate adiabatically connected to |10⟩. Also integrate the full time-dependent Lindblad equation along the CW and CCW paths and compare the state at E (or B) with the instantaneous steady state at the corresponding parameters. If F10 ≳ 0.99, the assumption holds; if F10 ≲ 0.95, the finite-detuning admixture is the main fidelity bottleneck and the mechanism description should be revised.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The protocol's physical mechanism, described after Fig. 1(b), requires that at the loop corners B and D (δ=±Δδ) the steady state of the Liouvillian L in Eq. (2) is 'very close to |10⟩' and that |10⟩ is 'also close to the eigenstate of H0' adiabatically connected to the target Bell state. However, for finite δ, |10⟩ is not an eigenstate of H0: from Eq. (1), H0|10⟩ = ξ|10⟩ + g|01⟩, so the coherent coupling g gives L(|10⟩⟨10|) ≠ 0. The steady state therefore necessarily contains admixtures of |01⟩, |00⟩, and |11⟩. The chosen parameters (Δδ=0.04, g=0.01) give δ/g = 4, which only modestly satisfies 'sufficiently large Δδ'. The paper does not report the numerical overlap of the steady state with |10⟩ at the corners, nor with the relevant H0 eigenstate. Since the final Bell-state fidelity cannot exceed the fidelity of the state at the start of the Hermitian segment with the corresponding eigenstate, this unquantified overlap is a genuine bottleneck: if it is ~0.95, the observed fidelity 0.9336 is fully explained by this effect alone, making the mechanism's central assumption the limiting factor rather than a robust high-fidelity preparation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript reports an experimental and numerical study of chirality-dependent entangled-state preparation in a dissipative two-qubit photonic system. The protocol combines steady-state engineering via a Lindblad jump operator with adiabatic passage along a closed loop in the (γ, δ) parameter space. The authors claim that the final Bell state (|Ψ+2⟩ or |Ψ−2⟩) is determined solely by the encircling direction, with experimental fidelities of 0.9336 and 0.9253, and that the scheme extends to three-qubit GHZ states. The dynamics are implemented experimentally using a quantum Langevin equation and a photonic platform that simulates the Liouvillian through general non-unitary evolutions.","tokens_in":12053,"tokens_out":25547,"duration_ms":213623,"significance":"If substantiated, the work would be a valuable contribution to dissipative state preparation and chiral quantum dynamics, demonstrating that path-dependent steady-state selection can produce entangled states with high fidelity. The photonic implementation of the full Liouvillian without post-selection is an experimental strength, and the extension to multipartite GHZ states is a useful step. However, the theoretical explanation of the mechanism contains a sign/overlap inconsistency that directly affects the central claim, and the quantitative verification of the key overlap assumption is missing. These issues are correctable but require careful revision.","major_comments":[{"comment":"The text states that at point E (δ = −Δδ for the CW path, per Appendix A) the steady state is very close to |10⟩, and that |10⟩ is also close to the eigenstate of H0 adiabatically connected to |Ψ+2⟩. This is inconsistent with Eq. (1). For the experimental parameters (Δδ = 0.04, g = 0.01), at δ = −0.04 the H0 eigenstate connected to |Ψ+2⟩ at δ = 0 is predominantly |01⟩, with |⟨10|v+⟩|² ≈ 0.05; the eigenstate with large |10⟩ overlap is the lower branch, which connects to |Ψ−2⟩ at δ = 0. Thus, if the steady state were close to |10⟩, CW encircling would produce |Ψ−2⟩, not |Ψ+2⟩. The authors must correct this sign/overlap inconsistency and confirm which eigenstate is actually populated at the corners.","section":"General mechanism, paragraph after Fig. 1(b)"},{"comment":"The claim 'For sufficiently large Δδ, the steady state is very close to |10⟩' is not quantified. With Δδ = 0.04 and g = 0.01, δ/g = 4, which only modestly satisfies the large-detuning condition. Since H0|10⟩ = ξ|10⟩ + g|01⟩, the steady state of L in Eq. (2) necessarily contains an admixture of |01⟩, and the final Bell-state fidelity cannot exceed the fidelity of the state at the start of the Hermitian segment with the corresponding H0 eigenstate. The paper does not report this overlap. The authors should provide the numerically computed overlap of the steady state with |10⟩ (and with the relevant eigenstate) at the corners, as a function of Δδ/g, and show that the working point lies in the high-overlap regime. Without this, the observed fidelities ~0.93 cannot be attributed to the proposed mechanism.","section":"Appendix A and the mechanism paragraph"},{"comment":"The final density matrices are obtained by averaging over n = 10 independent noise realizations of the quantum Langevin equation. The quoted error bars (e.g., F₂⁺ = 0.9336 ± 0.0005) are statistical deviations from Monte Carlo photon-counting statistics and do not include the spread over the 10 trajectories. With only 10 realizations, the sampling error of the ensemble average can be substantial. The manuscript should report the mean and standard deviation across the trajectories (or use bootstrap resampling), or increase n, before claiming high precision for the reported fidelities and concurrences.","section":"Eq. (5) and Fig. 2"}],"minor_comments":[{"comment":"The red and blue curves in the upper inset are referenced in the text but are not explicitly labeled in the figure caption. Please add labels or a full caption describing which eigenstate branch corresponds to |Ψ+2⟩ and |Ψ−2⟩.","section":"Fig. 1(b) inset"},{"comment":"The Conclusion states that the scheme 'does not involve post selection', but Appendix D discusses a post-selection strength α. The distinction between the full Liouvillian dynamics (α = 1, used in the main experiment) and the no-click limit (α = 0) should be clarified to avoid confusion.","section":"Appendix D and Conclusion"},{"comment":"The protocol relies on adiabatic following along the Hermitian sector, but no adiabaticity criterion or estimate of non-adiabatic transitions is given. A brief analysis of the Landau-Zener-type probability for the E→A and B→A segments with T = 1500 would strengthen the mechanism discussion.","section":"Adiabatic condition"},{"comment":"The dephasing strength (0.01) is small compared to γ = 0.18, so the robustness claim is rather weak. Showing the fidelity as a function of dephasing rate would make the claim more convincing.","section":"Robustness analysis"}],"recommendation":"major_revision","confidential_remarks":"The sign inconsistency in the mechanism description is a serious concern: if the reported parameterization and Bell-state definitions are correct, the text's explanation predicts the opposite chirality. This should be resolved before publication. The missing steady-state overlap analysis is also important because it is the central assumption of the protocol. I recommend requesting a corrected explanation with quantitative overlap data and a more careful statistical treatment of the trajectory ensemble."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nRead this one because of the title. It does what it says: for a two-qubit system with engineered dissipation, a slow closed loop in the (γ, δ) plane sends the state to |Ψ+2⟩ going clockwise and |Ψ−2⟩ going counterclockwise, with measured fidelities 0.9336 and 0.9253 and concurrence above 0.85. They also show a three-qubit GHZ version with fidelities around 0.92. The photonic implementation of the Liouvillian via the quantum Langevin equation is nontrivial, and the decomposition of the non-unitary evolution into beam displacers and wave plates looks like a real technical step. Robustness to dephasing and to random Hamiltonian perturbations is shown, and the protocol works without postselection—that last point separates it from the chiral state-transfer literature around exceptional points.\n\nWhat's new: prior chiral state transfer moved single states; this is the first (to my eye) use of loop chirality to select between two entangled steady states. The combination of dissipative steady-state engineering on one segment and adiabatic passage on the Hermitian segment is the right idea.\n\nSoft spots.\n\nThe mechanism's central assumption is stated, then not checked. The paper says that for large Δδ the steady state at the corners is 'very close to |10⟩', and |10⟩ is close to the H0 eigenstate that connects to the target Bell state. But at the experimental parameters Δδ=0.04, g=0.01, you have δ/g=4. Since H0|10⟩ = ξ|10⟩ + g|01⟩, the coherent coupling is one quarter of the detuning. The steady state will have a non-negligible |01⟩ admixture, and the overlap with |10⟩ is probably around 0.94–0.95. That is almost exactly the fidelity they measure. So the finite-detuning overlap is likely the limiting factor, not experimental noise. The authors should report the numerical overlap at the corners and, ideally, show how fidelity scales with δ/g. This is a missing check, not a refutation—the chirality still works, and the measured states are entangled. But the 'very close' claim is doing work and it is unquantified.\n\nThe robustness section uses only 10 noise realizations, and the quoted error bars are from Poissonian photon statistics, not the realization-to-realization spread. The shaded regions in the figures give a sense of the range, but no number is reported for the average or standard deviation over realizations. That is a minor reporting issue.\n\nThe scalability claim rests on a single three-qubit example, and the three-qubit Hamiltonian is not the direct N-qubit generalization of the two-qubit one—it's an Ising-type model with dissipation. Fine as a demonstration, but 'scalable' is an overstatement.\n\nThis paper is for people working on dissipative state preparation and photonic simulations of open quantum systems. The central result is new and supported by the data. The missing overlap check is worth asking for, but it does not sink the paper. I would send it to a serious referee.","headline":"Chiral dissipative entanglement generation is real and well demonstrated in two qubits; the main gap is an unquantified steady-state overlap at the loop corners, not a broken mechanism.","tokens_in":12445,"tokens_out":5786,"would_cite":true,"duration_ms":56049,"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":"A slow closed loop in parameter space turns dissipation into a directional switch that chooses which entangled state is prepared, with the direction of travel selecting the Bell state.","keywords":["chiral dynamics","dissipative entanglement generation","Bell states","GHZ states","Liouvillian dynamics","adiabatic passage","quantum Langevin equation","reservoir engineering"],"falsifier":"Run the loop at increasing speed or with a smaller detuning excursion: if the adiabatic-following premise fails, the final fidelity should drop and the two loop directions should no longer yield orthogonal Bell states; the crossover speed or the critical detuning range would be a direct, observable signature.","tokens_in":11615,"feed_emoji":"🔁","tokens_out":5265,"duration_ms":53947,"temperature":0.7,"pith_summary":"The paper seeks to establish that dissipation, normally an enemy of entanglement, can be engineered so that the direction in which system parameters are slowly varied around a closed loop decides which entangled state is produced. The central demonstration is a two-qubit photon system where clockwise loops prepare one Bell state and counterclockwise loops prepare the orthogonal one, with measured fidelities around 0.93. The same construction is shown to work for three-qubit GHZ states, with fidelities above 0.91, and to tolerate dephasing and random perturbations. The significance is that the final state is selected geometrically by the loop's chirality rather than by the initial state or by post-selection, offering a scalable route to controllable multipartite entanglement.","feed_headline":"Loop direction decides which entangled state emerges","feed_subtitle":"Dissipative two-qubit dynamics steers photons to orthogonal Bell states, then extends to GHZ states, with fidelities above 0.91.","key_machinery":"The engine is a designed non-unitary dynamics described by a Lindblad master equation whose Liouvillian (the generator of the dissipative evolution) has a dark state |10⟩ at large detuning and whose Hermitian sector has Bell states as eigenstates when the detuning vanishes. Traversing a slow closed loop in the (γ, δ) plane combines dissipative relaxation on the legs where the Liouvillian gap is large with adiabatic following on the leg where dynamics is Hermitian, so the state is deterministically transferred from one instantaneous steady state to another. The experiment implements the equivalent quantum Langevin equation stroboscopically on photon polarizations, allowing a general non-unita","core_discovery":"Starting from a maximally mixed state, encircling a closed rectangle in the (γ, δ) parameter plane one way drives a two-qubit system to the Bell state |Ψ+2⟩, while encircling it the opposite way drives it to the orthogonal Bell state |Ψ−2⟩. The measured final fidelities are 0.9336 and 0.9253, with concurrence near 0.86. The same loop protocol, applied to a three-qubit Hamiltonian, prepares the GHZ-type states |Ψ+3⟩ and |Ψ−3⟩ with fidelities above 0.91. The mechanism combines dissipative steady-state engineering – a large Liouvillian gap on the dissipative segments forces the system toward a dark state – with adiabatic passage on the Hermitian segment, so the system is handed from one instant","pith_inferences":["A natural extension is to probe whether loops with different winding numbers or non-rectangular shapes select other states in the same manifold; the paper's overlap conditions suggest that only specially designed loops will work, which is a testable prediction.","Because the final state is determined by loop direction rather than by the initial state or by post-selection, the protocol could serve as a passive directional state-preparation primitive in larger quantum information tasks, such as a reset operation that prepares a known entangled resource.","The mechanism's resilience to dephasing hints that this dissipative geometry could also be used to stabilize entanglement against slow parameter drifts, perhaps by repeating the loop or by using the loop orientation as a feedback variable.","A straightforward scaling test would apply the same loop design to four or more qubits; if the three-qubit generalization is representative, high fidelity should persist, though the required parameter ranges and Liouvillian gaps may need rebalancing."],"forward_implications":["A slow closed loop in parameter space prepares different Bell states depending solely on the loop's orientation, starting from a maximally mixed state and without post-selection.","The same loop design, with a different Hamiltonian and jump operators, prepares three-qubit GHZ-type states with fidelities above 0.91.","The prepared entanglement tolerates moderate dephasing and random Hamiltonian perturbations, with only slight reductions in fidelity and concurrence.","The chiral effect does not depend on exceptional-point encircling; it persists across the whole hybrid-Liouvillian family from full Liouvillian dynamics to the no-click non-Hermitian limit.","The scheme is explicitly extendable to multipartite entanglement, making it a candidate tool for scalable state preparation in open quantum systems."],"fun_headline_variants":["Chirality decides which Bell state emerges","Loop direction selects entangled state in dissipative dynamics","Dissipative path chirality picks Bell or GHZ state","Chiral loops create chosen entangled states","Which way you loop determines the entangled state"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The plan works only if the parameters are varied slowly enough for the system to keep up with the instantaneous steady state on the dissipative parts of the loop and the instantaneous eigenstate on the Hermitian part, and if the steady state at large detuning really is close to the dark state |10⟩.","fun_headline_variants_meta":{"raw":{"variants":["Chirality decides which Bell state emerges","Loop direction selects entangled state in dissipative dynamics","Dissipative path chirality picks Bell or GHZ state","Chiral loops create chosen entangled states","Which way you loop determines the entangled state"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000494,"raw_usage":{"total_tokens":2243,"prompt_tokens":709,"completion_tokens":1534,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":453,"completion_tokens_details":{"reasoning_tokens":1474}},"tokens_in":453,"tokens_out":1534,"duration_ms":11001,"temperature":1.0,"reasoning_tokens":1474,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T18:23:56.026760+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the loop at increasing speed or with a smaller detuning excursion: if the adiabatic-following premise fails, the final fidelity should drop and the two loop directions should no longer yield orthogonal Bell states; the crossover speed or the critical detuning range would be a direct, observable signature.","supporting_citations":[],"review_version":1}