{"id":"f1a0c005-b51f-468e-b7a9-301d85dd3a32","arxiv_id":"2509.15065","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"The Gaussian vacuum-projection steps in a standard CV entanglement distillation protocol can be absorbed into the preparation of the input Gaussian states, leaving an equivalent protocol that uses only single-photon detections.","lead":"A theory paper shows that continuous-variable entanglement distillation can be simplified by removing all Gaussian measurements, replacing them with specially prepared Gaussian input states, so that only single-photon detections are needed. The simplified protocol produces identical outputs with a higher success probability and clarifies how Gaussification can emerge without any Gaussian measurements.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The beam-splitter reordering identity (Eq. 7) is false as printed, so the central equivalence between the original and simplified distillation circuits is not established.","rationale":"The reader's weakest assumption is precisely Eq. (7), the unproved beam-splitter identity. My stress test agrees with that choice and strengthens it: the identity is not merely unproved but false under the paper's stated operator definitions, so the derivation of the simplified equivalent circuit is invalid at its key reordering step. This is the single most load-bearing point because the title and abstract claim 'completely equivalent outputs', and that equivalence is what justifies replacing the original heralded-Gaussification scheme with the simplified one. The multicopy version inherits the same problem because its simplification is described as following the same steps. I do not claim the final polynomial form or the asymptotic Gaussification result is necessarily wrong; it is possible that a corrected identity or a direct calculation of the simplified network could salvage the equivalence. But as the preprint stands, the central equivalence is unsupported by a demonstrably incorrect intermediate step. This leaves the reader's CONDITIONAL verdict appropriate: the author should provide a corrected identity or a direct derivation that does not rely on Eq. (7), and should display the calculation connecting Eq. (18) to Eq. (13). Since my concern does not establish a final negative result, I recommend no change to the verdict; the paper should remain conditional pending those checks.","tokens_in":21167,"tokens_out":58684,"duration_ms":500203,"concrete_test":"Compute the two 4×4 passive-mode transformation matrices L=R12(π/4)R13(θ)R24(θ)R34(−π/4) and R=R34(−π/4)R13(θ)R24(θ)R12(π/4) using the definitions (3) and (8), with T=0.8, and compare L and R on the basis vectors |1,0,0,0⟩ and |0,1,0,0⟩; if L≠R, Eq. (7) fails as written. Independently, recompute the normalized output state from Eq. (18) for κ=1 and compare the coefficient ratios with Eq. (13); if the ratios differ for n=0 and n=1, the claimed equivalence between the simplified and original outputs is not confirmed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The equivalence chain from Fig. 2(a) to Fig. 2(e) hinges on Eq. (7), which is used to justify the reordering step Fig. 2(b) to Fig. 2(c). With the paper's own definitions (3) and (8), U_{A1A2}=exp[π/4(a1†a2−a1a2†)] and V_{AjCj}=exp[θ(aj†cj−ajcj†)] with θ=arccos√T, the identity as stated does not hold. A direct evaluation on the four-mode Fock state |1,0,0,0⟩ at T=0.8 gives different mode amplitudes on the two sides: the left-hand side has no component in the C2 mode while the right-hand side does. Hence the reordering step is invalid, and the subsequent absorption of the Gaussian vacuum projections into the prepared state σ is unsupported. The same issue propagates to the multicopy construction in Sec. VII, which is said to proceed 'in the same way' as the two-copy case. The reader's concern about Eq. (7) is therefore not merely that a proof is missing; the printed identity appears to be wrong under the stated conventions. A separate asserted equality, that Eq. (18)/(19) reproduces Eq. (13), should also be checked explicitly, since it is another unshown algebraic step in the equivalence claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a simplified continuous-variable entanglement distillation scheme in which the heralding Gaussian measurements (vacuum projections) of the original iterative Gaussification protocol are eliminated and absorbed into the preparation of modified Gaussian input states. The authors claim that the simplified circuit produces exactly the same output as the original protocol while requiring fewer detectors and having a higher success probability. They analyze the two-copy case for pure and mixed inputs, compare with generalized photon subtraction, and present a multicopy generalization in which the output converges to a two-mode squeezed vacuum with squeezing parameter 2Tλ. The material closely follows the recent simplification of GKP state breeding by Aghaee Rad et al.","tokens_in":111,"tokens_out":49598,"duration_ms":1273647,"significance":"If the claimed equivalence is correct, the result is significant: it would show that multicopy distillation of Gaussian entanglement can be achieved without any heralding Gaussian measurements, using only single-photon projections, and it would provide a bipartite analogue of the recent GKP breeding simplification. The paper also gives explicit formulas for the distilled squeezing variance, entanglement, and fidelity, and treats mixed states with loss, which are valuable for potential experiments. The multicopy output formula Eq. (49) and the interpretation as bipartite Gaussian boson sampling are attractive. However, the central equivalence is not established by the arguments in the manuscript: the key reordering identity is false as stated, and the two-copy derivation contains an algebraic error. The manuscript would require a substantially corrected proof before its main claim is reliable.","major_comments":[{"comment":"The beam-splitter reordering identity Eq. (7) is false as written. With the paper's definitions (3) and (8), U_{A1A2}=exp[π/4(a1†a2−a1a2†)] and V_{AjCj}=exp[θ(aj†cj−ajcj†)], θ=arccos√T, the two sides of Eq. (7) do not agree for generic T. A direct evaluation on |1,0,0,0⟩ at T=0.8 gives different coefficients in the C2 mode on the two sides. Since this identity is the sole justification for the reordering step from Fig. 2(b) to Fig. 2(c), the subsequent absorption of the Gaussian vacuum projections into the prepared state σ is unsupported. This is a load-bearing point for the claimed equivalence.","section":"§III, Eq. (7)"},{"comment":"The assertion that Eq. (18)/(19) reproduces Eq. (13) is incorrect. Inserting ν=(1−T)λ into Eq. (19), the generated state has Fock-space coefficients proportional to 2λ²µⁿ + n(n−1)µⁿ⁻² (up to a common prefactor), whereas Eq. (13) has coefficients proportional to (n²+3n+4)µⁿ. These are not proportional for general λ and T, so the simplified two-copy output as written is not equal to the original output. This error directly contradicts the claim that the two schemes produce 'completely equivalent outputs'.","section":"§IV, Eqs. (18)-(19)"},{"comment":"The multicopy simplification in Sec. VII is said to proceed 'in the same way' as the two-copy case, so it inherits the problems with Eq. (7). The reordering of the whole beam-splitter arrays relies on the same false identity, and therefore the proof of equivalence between Fig. 8(a) and Fig. 8(b) is not established. The output formula Eq. (49) may be correct—a direct M=2 calculation from the simplified circuit in Fig. 8(b) yields Eq. (49) and matches Eq. (13)—but the manuscript needs a correct derivation of this equivalence rather than the flawed commutation argument.","section":"§VII"}],"minor_comments":[{"comment":"There are numerous typos and small language errors, including 'suqeezed' (Sec. II), 'previouss' (Sec. IV), 'multicoppy' (Sec. VI), 'Gausisan' (Sec. VII), 'auxliary' (Introduction), and 'muticopy' (Sec. VIII).","section":"General"},{"comment":"The reasoning leading to Eq. (17) is unclear: the stated projection (|0,2>−|2,0>)/√2 has zero overlap with a two-mode squeezed vacuum state, so the conclusion that the ancilla modes are projected onto ν²|0,0>+|2,2> does not follow from the preceding sentence. This passage should be rewritten with a clear derivation.","section":"§IV, Eq. (17)"},{"comment":"The two roots for κ² should be discussed more explicitly: the text notes that κ² can be negative, but it would help to state which root is physically admissible and how the constraint |κ|<1/(1−T) is enforced in the optimization.","section":"§IV, Eq. (24)"},{"comment":"The caption states that 'both ρ and σ in panels (a)-(e) denote Gaussian states', while panel (f) refers to ρ^(i) as a non-Gaussian state; the notation should be clarified to avoid confusion about what is input to each stage.","section":"Fig. 2 caption"}],"recommendation":"major_revision","confidential_remarks":"The central equivalence claim is not proven as written, and I have identified explicit numerical counterexamples to Eq. (7) and a mismatch between Eq. (18)/(19) and Eq. (13). However, my direct calculation of the M=2 multicopy simplified circuit confirms that Eq. (49) reproduces the original protocol output, so the core idea may well be correct. The authors should be given the opportunity to replace the flawed commutation proof with a correct derivation and to fix the two-copy formulas. This is not a rejection of the underlying physics, but the manuscript in its current form is not publishable without these corrections."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe headline is this: the paper's central equivalence between the original heralded-Gaussification distillation circuit and the simplified no-Gaussian-measurement version rests on Eq. (7), a beam-splitter reordering identity that is stated without proof and appears to be false as printed. I checked the identity on the four-mode state |1,0,0,0> using the paper's own definitions (3) and (8) at T=0.8, and the two sides give different states: the left side couples excitation into C2, the right into C1. So the simplification chain from Fig. 2(a) to 2(e) is unsupported, and the multicopy version in Sec. VII inherits the problem.\n\nThat said, the paper is not careless. It explicitly credits Ref. [49] for the measurement-elimination trick, which is exactly right, and the extension to two-mode Gaussification is a natural and potentially useful step. The explicit output formulas (13), (20), (49), and the limit (1+µa†b†/M)^M → exp(µa†b†) are clean and likely correct as algebra. The mixed-state analysis in Sec. VI and the success-probability gain Pσ^{-(M-1)} are worked out carefully. If the reordering identity can be fixed to a correct version, the paper would make a solid contribution.\n\nThe main soft spot is exactly Eq. (7). The reader's report flagged it as missing proof; the stress test goes further and I agree with the stress test: the printed identity fails a direct Fock-state check. A second soft spot, minor by comparison, is that the claimed equality between Eqs. (18)/(19) and (13) is asserted without displaying the calculation; that one is probably true but should be shown. The paper's own caveats in Sec. VIII, simultaneous single-photon detections and PNR detectors, are stated honestly and don't affect the formal equivalence argument.\n\nCitation pattern looks fine; the self-citations are to the relevant Gaussification and noiseless-attenuation literature.\n\nRecommendation: send to peer review, but with a clear request to prove or correct Eq. (7) and to display the verification of Eq. (18)/(19) vs (13). As it stands, the central claim is not established, but this looks like a fixable technical error rather than a fundamentally wrong idea. I would not cite the current version in my own work until the identity is repaired.","headline":"The simplification rests on a beam-splitter identity that appears false as printed, so the main claim is unproven; still worth refereeing because the error looks fixable.","tokens_in":21950,"tokens_out":5743,"would_cite":false,"duration_ms":41663,"reading_group":"no","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["03.67.Bg","42.50.Dv"],"model":"deepseek-v4-flash","headline":"Gaussian vacuum projections in continuous-variable entanglement distillation can be absorbed into the preparation of input squeezed states, leaving only single-photon heralding with identical outputs, fewer detectors, and higher success…","keywords":["continuous-variable entanglement distillation","Gaussian states","single-photon subtraction","heralded Gaussification","two-mode squeezed vacuum","noiseless attenuation","Gaussian measurements","Gaussian boson sampling"],"falsifier":"Evaluate both sides of Eq. (7) exactly on a set of Fock states using the stated transmittance $T$ for the unbalanced beam splitters; if the operator identity fails for any $T$, the simplified circuit does not reproduce the original output. A direct experiment could run the original and simplified two-copy circuits at fixed $\\lambda$ and $T$ and compare the squeezing variances and success probabilities of the heralded outputs.","tokens_in":20900,"feed_emoji":"🔗","tokens_out":16763,"duration_ms":130708,"temperature":0.7,"pith_summary":"This paper asks whether the Gaussian measurements in a standard continuous-variable entanglement distillation protocol are genuinely necessary. Its answer is no: the projections onto vacuum used in the repeated Gaussification step can be commuted backwards through the circuit and absorbed into the preparation of the input squeezed states. The resulting simplified scheme uses only projections onto single-photon states, needs fewer detectors, and succeeds with higher probability while producing exactly the same outputs. For $M$ pure copies the output is $|\\Psi_{\\rm out}\\rangle \\propto (1+\\mu \\hat{a}^{\\dagger}\\hat{b}^{\\dagger}/M)^{M} e^{\\mu \\hat{a}^{\\dagger}\\hat{b}^{\\dagger}}|\\mathrm{vac}\\rangle$, which as $M\\to\\infty$ approaches $e^{2\\mu\\hat{a}^{\\dagger}\\hat{b}^{\\dagger}}|\\mathrm{vac}\\rangle$, a two-mode squeezed vacuum (the standard Gaussian entangled state of two light modes) with squeezing parameter $2T\\lambda$. If this is right, multicopy distillation of Gaussian entanglement is possible without any Gaussian measurement in the loop.","feed_headline":"Multicopy entanglement distillation needs no Gaussian measurements","feed_subtitle":"Gaussian measurements are replaced by single-photon clicks; the distilled state is identical, with fewer detectors.","key_machinery":"The argument runs on three operations. First, a passive beam-splitter reordering identity (Eq. 7), $\\hat{U}_{A_1A_2}\\hat{V}_{A_1C_1}\\hat{V}_{A_2C_2}\\hat{U}^{\\dagger}_{C_1C_2} = \\hat{U}^{\\dagger}_{C_1C_2}\\hat{V}_{A_1C_1}\\hat{V}_{A_2C_2}\\hat{U}_{A_1A_2}$, with $\\hat{U}$ a balanced (50:50) beam splitter and $\\hat{V}$ the unbalanced beam splitter used for photon subtraction, lets the Gaussian measurements move past the photon-subtraction splitters. Second, because two identical zero-displacement Gaussian states are left unchanged by balanced beam-splitter coupling (Eq. 9), the inner interferometers of the multicopy circuit can be removed. Third, the projection onto vacuum is reinterpreted as a conditional zero-photon subtraction, i.e. a noiseless attenuation of each mode, which converts the auxiliary input into $\\hat{\\sigma}\\propto (1-T)^{(\\hat{n}_A+\\hat{n}_B)/2}\\hat{\\rho}(1-T)^{(\\hat{n}_A+\\hat{n}_B)/2}$. In the $M$-copy version these steps turn the whole protocol into two $M$-mode interferometers whose first input is $\\hat{\\rho}$, whose remaining inputs are $\\hat{\\sigma}$, and whose auxiliary outputs are all projected onto $|1\\rangle$.","core_discovery":"The central discovery is a circuit equivalence: every Gaussian measurement in the heralded Gaussification stage can be removed. The original round combines local single-photon subtractions with balanced beam-splitter interference and projections of auxiliary modes onto vacuum; the simplified version absorbs those vacuum projections into a different input Gaussian state $\\hat{\\sigma}$, obtained by two-mode noiseless attenuation of the original state, and keeps only projections onto the single-photon state $|1\\rangle$. For pure two-mode squeezed vacuum inputs with squeezing parameter $\\lambda$ and photon-subtraction transmittance $T$, the two-copy output is proportional to $\\sum_{n=0}^{\\infty}(n^2+3n+4)(T\\lambda)^n |n,n\\rangle$, and for $M$ copies the output is proportional to $(1+\\mu \\hat{a}^{\\dagger}\\hat{b}^{\\dagger}/M)^M e^{\\mu\\hat{a}^{\\dagger}\\hat{b}^{\\dagger}}|\\mathrm{vac}\\rangle$ with $\\mu=T\\lambda$. In the large-$M$ limit this becomes a Gaussian two-mode squeezed vacuum with squeezing parameter $2T\\lambda$, so multicopy distillation converges to a Gaussian entangled state with no heralding Gaussian measurement. The same structure persists for mixed states distributed through lossy channels, where $\\hat{\\sigma}$ is a pumped-down squeezed thermal state and the local-operations-and-classical-communication structure of the distillation protocol is unchanged.","pith_inferences":["The same absorption trick should transfer to single-mode Gaussification and squeezing-distillation protocols, which share the parallel Gaussian-postselection structure; the paper demonstrates the two-mode case and cites the analogous single-mode logical-qubit breeding case.","Because the multicopy output factorizes as one factor per heralded pair of modes, the protocol is a bipartite analogue of Gaussian boson sampling; quantifying how photon-number-resolving detector inefficiency degrades the distilled squeezing is a natural next step.","A global optimization of all $M$ input squeezing parameters, rather than the single parameter $\\kappa$ scanned here, could trade success probability against distilled squeezing in ways the one-round analysis does not capture."],"forward_implications":["For $M$ pure copies the output approaches a Gaussian two-mode squeezed vacuum with squeezing parameter $2T\\lambda$ as $M\\to\\infty$, so Gaussian entanglement can be distilled without any Gaussian measurement.","The success probability is higher by a factor $1/P_\\sigma$ for every additional copy, giving an exponential improvement $P_\\sigma^{1-M}$ over the original multicopy scheme.","The auxiliary Gaussian states $\\hat{\\sigma}$ have weaker squeezing than the original states and can be generated deterministically from the same source, so the simplification does not demand a new non-Gaussian resource.","For mixed inputs distributed over lossy channels the distilled squeezing obeys a bound set by the loss, and the local-operations-and-classical-communication structure of the protocol is preserved, so the simplification is not limited to ideal pure-state operation."],"supporting_citations":[{"why":"Supplies the circuit-simplification strategy of absorbing Gaussian measurements into input-state preparation, adapted here from single-mode logical-qubit breeding to entanglement distillation.","marker":"[49]"},{"why":"Defines the iterative heralded Gaussification map and the convergence of photon-subtracted two-mode squeezed vacuum to a Gaussian state with squeezing $2T\\lambda$.","marker":"[41]"},{"why":"Establishes that linear optics and postselection can drive non-Gaussian states to Gaussian states, the mechanism being simplified.","marker":"[40]"},{"why":"Reports experimental distillation and purification of continuous-variable entangled states, the target capability.","marker":"[44]"},{"why":"Introduces the local single-photon subtraction operation used to de-Gaussify the input states.","marker":"[58]"},{"why":"Provides the noiseless-attenuation or conditional-zero-photon-subtraction formalism used to define the auxiliary state $\\hat{\\sigma}$.","marker":"[77]"},{"why":"Proposes the two-mode generalized photon subtraction scheme with an auxiliary Gaussian state that the simplified scheme is compared against in Sec. V.","marker":"[66]"},{"why":"Derives the equivalent circuits for preparing $\\hat{\\sigma}$ from lossy Gaussian states, used in the mixed-state analysis.","marker":"[76]"}],"fun_headline_variants":["Gaussian measurements vanish from entanglement distillation","Fewer detectors, same distilled entanglement","Single-photon clicks replace Gaussian measurements","Simplified distillation: no Gaussian measurements","Multicopy distillation skips Gaussian measurements"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The simplification stands or falls on the beam-splitter reordering identity of Eq. (7), which is stated without proof, together with the requirement that all Gaussian and non-Gaussian measurements happen in parallel so that their order can be swapped.","fun_headline_variants_meta":{"raw":{"variants":["Gaussian measurements vanish from entanglement distillation","Fewer detectors, same distilled entanglement","Single-photon clicks replace Gaussian measurements","Simplified distillation: no Gaussian measurements","Multicopy distillation skips Gaussian measurements"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000288,"raw_usage":{"total_tokens":1715,"prompt_tokens":993,"completion_tokens":722,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":609,"completion_tokens_details":{"reasoning_tokens":660}},"tokens_in":609,"tokens_out":722,"duration_ms":6431,"temperature":1.0,"reasoning_tokens":660,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T15:51:13.942072+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Evaluate both sides of Eq. (7) exactly on a set of Fock states using the stated transmittance $T$ for the unbalanced beam splitters; if the operator identity fails for any $T$, the simplified circuit does not reproduce the original output. A direct experiment could run the original and simplified two-copy circuits at fixed $\\lambda$ and $T$ and compare the squeezing variances and success probabilities of the heralded outputs.","supporting_citations":[{"cited_title":"Franzen, B","cited_arxiv_id":null,"evidence_quote":"Supplies the circuit-simplification strategy of absorbing Gaussian measurements into input-state preparation, adapted here from single-mode logical-qubit breeding to entanglement distillation."},{"cited_title":"Encoding a qubit in an oscillator","cited_arxiv_id":null,"evidence_quote":"Defines the iterative heralded Gaussification map and the convergence of photon-subtracted two-mode squeezed vacuum to a Gaussian state with squeezing $2T\\lambda$."},{"cited_title":"Ourjoumtsev, H","cited_arxiv_id":null,"evidence_quote":"Establishes that linear optics and postselection can drive non-Gaussian states to Gaussian states, the mechanism being simplified."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Reports experimental distillation and purification of continuous-variable entangled states, the target capability."},{"cited_title":"Garc´ ıa-Patr´ on, J","cited_arxiv_id":null,"evidence_quote":"Provides the noiseless-attenuation or conditional-zero-photon-subtraction formalism used to define the auxiliary state $\\hat{\\sigma}$."},{"cited_title":"Giedke and J","cited_arxiv_id":null,"evidence_quote":"Proposes the two-mode generalized photon subtraction scheme with an auxiliary Gaussian state that the simplified scheme is compared against in Sec. V."},{"cited_title":"Nha and H","cited_arxiv_id":null,"evidence_quote":"Derives the equivalent circuits for preparing $\\hat{\\sigma}$ from lossy Gaussian states, used in the mixed-state analysis."}],"review_version":2}