{"id":"30fdcb9c-315e-42ad-9e95-865c9dfb7acf","arxiv_id":"2608.06229","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"For KAGRA's post-O5 upgrade, a filter-cavity frequency-dependent squeezing scheme yields the best binary neutron star range, beating frequency-independent squeezing by at least 23% in detection rate, while EPR squeezing is best for heavy binaries.","lead":"This paper simulates five quantum noise reduction schemes for the KAGRA gravitational wave detector upgrade and compares their detection ranges. It concludes that a detuned filter cavity gives the best binary neutron star range once low-frequency noise is quantum-limited, and EPR squeezing is best for heavy binary systems.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Dephasing/locking noise is acknowledged but not modeled for the detuned 85 m filter cavity; the claimed 23% rate gain corresponds to only ~7% BNS-range advantage, so unmodeled phase noise could erase the FC recommendation.","rationale":"The paper is a competent design comparison for KAGRA post-O5, with transparent parameter tables, two classical-noise scenarios, and an explicit statement that filter-cavity parameters are optimized for BNS range rather than all sources. However, the strongest quantitative claim is specifically the FC advantage over FIS/EPR, and the authors themselves flag dephasing and locking precision as the main issue for short filter cavities. Because the claimed 23% rate gain is equivalent to only about 7% BNS-range gain, the omission of any dephasing model is not a minor technicality: even a few percent of sensitivity degradation could move the comparison. This is the single most load-bearing concern because it directly targets the recommendation to build a detuned 85 m filter cavity, and it is acknowledged but left unquantified in the manuscript. The reader's weakest-assumption analysis identifies the same issue, and my independent reading agrees. Other potential issues, such as the empirical loss scaling law or asymmetric optimization, are less decisive because the paper brackets losses between 13 ppm and 38 ppm and the main ranking is not shown to depend on the exact scaling. No machine-checked proofs, code, or data are provided, so numerical reproducibility is limited, but that is secondary to the physical omission. The conditional verdict is therefore appropriate; my stress-test does not move it.","tokens_in":13086,"tokens_out":10596,"duration_ms":107228,"concrete_test":"Recompute the Fig. 6 BNS range for the 85 m FC after injecting a dephasing/locking-error term into the filter-cavity transfer matrix, using the Kwee et al. (PRD 90, 062006) phase-noise/decoherence formalism. Use residual length noise equal to the measured LIGO filter-cavity value scaled to 85 m, and also determine the length-noise level at which the FC-vs-FIS range advantage falls below 7%. If realistic dephasing preserves more than 7% range gain at both 13 ppm and 38 ppm losses, the 23% claim stands; otherwise the FC recommendation and the rate-gain estimate should be downgraded.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central recommendation (FC scheme gives the largest BNS range and at least a 23% detection-rate increase) rests on an unmodeled dephasing/locking-error term. Section I explicitly states that dephasing and locking precision are the main issue when filter cavities become short (Ref. [37]), and the Fig. 5 discussion concedes that AFC and FDBS avoid the dephasing effect 'associated with the latter.' Yet the quantitative BNS-range comparison and the 23% detection-rate claim include no phase-noise or detuning-jitter contribution in the FC transfer matrix (Eqs. 14 and 16). The claimed rate gain is fragile: a 23% rate increase corresponds to only about 7% BNS-range improvement, since rate scales roughly as range cubed. If residual length noise or detuning jitter degrades the FC sensitivity by more than a few percent, the FC advantage over FIS or EPR could vanish or reverse. The paper itself identifies the missing physical ingredient but never bounds its size or shows that the conclusion is robust to it.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript compares five quantum noise reduction schemes—frequency-independent squeezing (FIS), filter-cavity frequency-dependent squeezing (FC), amplitude filter cavity (AFC), frequency-dependent beam splitter (FDBS), and EPR squeezing—for the KAGRA post-O5 high-frequency configuration. The authors develop a general cavity input-output model with loss ports, propose an empirical loss scaling law for filter cavities as a function of equivalent confocal length, and evaluate the astrophysical BNS range using an external detector model. They use particle swarm optimization to tune filter-cavity parameters and conclude that FC outperforms AFC and FDBS at all frequencies, that FC provides the largest BNS range when low-frequency noise is quantum dominated, and that an 85 m FC yields a 23–48% increase in BNS detection rate compared with FIS, with EPR best for heavy binaries.","tokens_in":13343,"tokens_out":4738,"duration_ms":46097,"significance":"The study is timely and practically relevant for KAGRA's post-O5 planning, and it performs a broader comparison of five quantum noise reduction schemes than is usually presented. The use of particle swarm optimization instead of the analytic filter-cavity formulas for loss-dominated cavities is a sensible methodological choice, and the use of an external detector model for BNS range evaluation is a strength. However, the central recommendation for FC is built on two load-bearing assumptions that are not fully quantified: the absence of dephasing and locking errors in the detuned FC, and the validity of the empirical loss scaling law. If the authors can model or bound the dephasing effect and provide a more complete characterization of the loss scaling, the work would provide a reliable quantitative basis for the KAGRA upgrade decision.","major_comments":[{"comment":"The text explicitly states that 'Neither the AFC nor the FDBS scheme outperforms the FC scheme, despite the absence of the dephasing effect associated with the latter,' and Section I cites Ref. [37] identifying dephasing and locking precision as the main issue for short filter cavities. Yet the FC transfer matrix in Eqs. 14 and 16 contains no phase-noise or detuning-jitter term, and the BNS-range comparison in Section V does not include one. Since a 23% detection-rate increase corresponds to roughly a 7% BNS-range improvement, an unmodeled few-percent degradation of FC sensitivity could erase or reverse the claimed advantage over FIS or EPR. The manuscript should model the dephasing/locking noise or bound its magnitude and demonstrate that the conclusion is robust.","section":"Section V, Fig. 5 discussion"},{"comment":"Equation (9), L = 10 x L_conf^0.3, is presented as an empirical scaling law for cavity losses, but the paper gives no underlying data table, no fit uncertainties, no number of data points, and no fit-quality measure. Figure 2 shows a blue curve and a few measured points, but the extrapolation to an 85 m cavity is load-bearing: Table I's optimal input-mirror transmissivities (58 and 63 ppm at 38 and 13 ppm losses) and the range curves in Fig. 6 depend directly on the assumed loss level. Please report the full set of measured loss values, the fitting procedure, and the sensitivity of the BNS-range ranking to the scaling coefficient and exponent.","section":"Section II C, Eq. (9)"},{"comment":"The particle swarm optimization is described as optimizing three parameters—squeezing level, input-mirror transmissivity, and detuning—but Table I reports only transmissivity values. The optimized detunings for the 40 m and 85 m cases, and the resulting filter-cavity bandwidths, are not reported anywhere, and Eqs. 24–25 are dismissed as no longer optimal without a replacement formula. Without the detuning values and the actual filter-cavity transfer functions used, the quantitative claims in Figs. 5–7 and the 23–48% rate gain cannot be independently reproduced or checked.","section":"Section IV, Table I"}],"minor_comments":[{"comment":"The captions contain the typo 'fucntion' for 'function', and the Introduction contains 'nuclesosynthesis' for 'nucleosynthesis'; these should be corrected during editing.","section":"Figure captions 2 and 3"},{"comment":"Equation (8) uses the symbol L both for round-trip loss and for cavity length, which makes the expression 'L/(2τ) = c L/(2 L)' confusing; please distinguish the two quantities with different symbols.","section":"Section II C, Eq. (8)"},{"comment":"The appendix on the low-quality-factor suspension scenario is very brief and contains no equations or description of how the low-quality suspension is modeled; it should either be expanded or integrated into the main text with a short description.","section":"Section VII, Appendix A"},{"comment":"Reference [44] contains LaTeX artifacts ('<? tex \\break?>') and several references have inconsistent formatting; the reference list should be cleaned before submission.","section":"References"},{"comment":"The abstract states 'at least 23% increase' without specifying the loss assumptions, while the text gives a range 23–48% depending on the 13 ppm and 38 ppm cases; please state the assumed loss conditions in the abstract for precision.","section":"Abstract and Section V"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is within the scope of the journal as a detector-characterization and noise-reduction study for KAGRA. The central risk is the unmodeled dephasing/locking noise of the detuned FC, which the authors themselves flag in Fig. 5's discussion; if they can add a quantitative dephasing model or show that the FC conclusion survives with a realistic dephasing budget, the paper would be a solid contribution. The empirical loss scaling law also needs more transparency. I would not reject the paper at this stage, but the missing quantitative robustness checks should be addressed in a revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a useful engineering comparison for KAGRA post-O5, not a fundamental advance. It applies known squeezing schemes to a fixed-bandwidth configuration, contributes an empirical loss scaling relation, and gives optimized filter cavity parameters. The claim that FC gives the largest BNS range is probably right within the model, but the headline 23% detection-rate gain rests on unmodeled dephasing and an asymmetric optimization. The paper itself acknowledges the dephasing issue but does not bound it.\n\nWhat's new: the specific KAGRA post-O5 HF configuration comparison, the PSO-optimized parameter table, and a rough empirical loss law L = 10 × L_conf^0.3 from a handful of measurements. These are legitimately useful. The cavity and two-photon formalisms are standard and internally consistent. Figures 5–7 give a clear picture of the trade-offs, and the observation that FC beats EPR for BNS but not for heavy binaries is a useful nuance.\n\nSoft spots: (1) Dephasing and locking errors are mentioned as the main issue for short filter cavities, and the AFC/FDBS discussion notes they avoid it, but the quantitative comparison and the 23% number include no phase-noise term. The FC advantage over FIS is only a few percent in range, so a modest phase noise could erase it. (2) The empirical loss scaling Eq. 9 is fit to very few points with no uncertainty; it drives the extrapolation to other cavity lengths and losses. This is weak but not fatal, since the results are bracketed by best-case and empirical cases. (3) The optimization appears asymmetric: FC parameters are optimized for BNS range, but it is not clear that EPR or FIS parameters are optimized to the same standard. That could bias the comparison. (4) No code or data are provided, so numerical results cannot be independently checked.\n\nWho this is for: KAGRA collaboration members and people designing squeezing upgrades for gravitational-wave detectors. It deserves a serious referee because the recommendation affects real hardware choices. The missing dephasing estimate is specific and addressable; the paper should either model it or explicitly argue it is negligible at the 85 m scale.","headline":"Useful, competent KAGRA design comparison, but the headline FC advantage rests on unmodeled dephasing and a low-data loss fit.","tokens_in":13820,"tokens_out":2229,"would_cite":false,"duration_ms":21494,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["04.80.Nn"],"model":"deepseek-v4-flash","headline":"For KAGRA post-O5, a detuned 85 m filter cavity with optimized parameters gives the greatest binary-neutron-star detection range among compared quantum-noise schemes, raising the detection rate by at least 23% over frequency-independent…","keywords":["gravitational wave detectors","quantum noise","squeezed vacuum","frequency-dependent squeezing","filter cavity","KAGRA","binary neutron star range","EPR squeezing"],"falsifier":"Lock an 85 m detuned, over-coupled filter cavity with the proposed parameters and measure the squeezing degradation due to dephasing and locking noise at the operating point; if the degradation is large enough that less than about 15 dB of effective squeezing reaches the interferometer, the predicted at least 23% BNS detection-rate gain over FIS will not appear in the measured sensitivity.","tokens_in":12879,"feed_emoji":"🔭","tokens_out":7728,"duration_ms":66297,"temperature":0.7,"pith_summary":"The paper asks which quantum-noise-reduction scheme KAGRA should adopt after its fifth observing run, given that the underground site allows only about 85 m for a filter cavity. Comparing frequency-independent squeezing (FIS), filter-cavity frequency-dependent squeezing (FC), an amplitude filter cavity (AFC), a frequency-dependent beam splitter (FDBS), and an EPR two-mode-squeezing scheme, it finds that FC outperforms AFC and FDBS at all frequencies. When low-frequency noise is quantum-dominated, an 85 m detuned filter cavity with numerically optimized input-mirror transmissivity and detuning provides the largest binary-neutron-star (BNS) range, at least 23% higher detection rate than FIS; when low-frequency noise is classical-dominated, FIS wins. EPR is best for heavier binary systems. The paper also argues that filter-cavity detuning alone can be re-tuned to compensate for arm-power and loss variations after construction.","feed_headline":"Detuned 85 m filter cavity boosts KAGRA BNS detections by 23%","feed_subtitle":"Optimized mirror transmission and detuning make frequency-dependent squeezing the best post-O5 choice.","key_machinery":"The central object is the detuned, over-coupled filter cavity, modeled by a frequency-dependent 2x2 transfer matrix that rotates the injected squeezed-vacuum quadrature from phase squeezing at high frequencies toward amplitude squeezing at low frequencies. The argument is carried by a three-port input-output model of the cavity with losses entering as additional vacuum ports, an empirical loss scaling law (loss per round trip proportional to the 0.3 power of the confocal cavity length) used to set realistic round-trip losses for the 85 m cavity, and particle-swarm optimization over squeezing level, input-mirror transmissivity, and detuning. The optimized parameters, capped at 15 dB squeezing, yield the claimed BNS range gains.","core_discovery":"On the paper's own terms, the central discovery is that for KAGRA post-O5's high-frequency configuration, the conventional detuned filter-cavity scheme is the best single-mode choice and beats the EPR scheme for BNS sources under the realistic 85 m length constraint, provided filter-cavity parameters are numerically optimized rather than set by analytic formulas. With 13 ppm round-trip losses and optimized input mirror transmissivity (about 63 ppm) and detuning, the FC scheme improves the BNS detection rate by at least 23% compared with frequency-independent squeezing; the paper also reports a 7% to 14% BNS-range gain over the EPR scheme, corresponding to a 23% to 48% detection-rate increase depending on loss assumptions. FIS remains preferred when low-frequency noise is dominated by classical noise, and EPR is preferred for heavy binary systems. AFC and FDBS schemes are not competitive.","pith_inferences":["If KAGRA's actual suspension quality lands between the high- and low-quality scenarios considered, the crossover between FIS and FC will shift; a measured sensitivity curve after upgrade could decide which scheme to activate.","The same parameter-optimization and loss-scaling approach could be applied to other space-constrained underground detectors to choose between a filter cavity and EPR squeezing.","The claim that detuning alone compensates for arm-power and loss variations assumes the detuning actuator has enough range and precision, which the paper does not quantify.","The 23% detection-rate gain assumes an isotropic BNS source distribution; a strongly anisotropic distribution could make the realized rate gain differ from the quoted figure."],"forward_implications":["KAGRA should implement a detuned 85 m filter cavity with optimized input-mirror transmissivity and detuning to maximize BNS detection rate under quantum-noise-dominated low frequencies.","If low-frequency noise remains classical-dominated, as with low-quality-factor suspensions, frequency-independent squeezing is the better and simpler choice, needing no filter cavity.","Filter-cavity detuning can be re-optimized after commissioning to compensate for arm-power changes between half and full design value and for different intra-cavity loss conditions, without replacing the input mirror.","AFC and FDBS schemes can be deprioritized for KAGRA post-O5, while the EPR scheme should be retained as the option for heavier binary systems.","The reported 7% to 14% BNS-range gain over EPR translates to 23% to 48% more detections, making the FC choice consequential for event rates."],"supporting_citations":[{"why":"Introduces the filter-cavity scheme for frequency-dependent squeezing that all FC comparisons build on.","marker":"[16]"},{"why":"Shows a single filter cavity suffices for RSE-configuration interferometers, justifying the 85 m single-cavity design.","marker":"[17]"},{"why":"Supplies the empirical loss scaling and confocal-length normalization used to set realistic round-trip losses.","marker":"[18]"},{"why":"Defines the amplitude filter cavity scheme that the paper compares against FC.","marker":"[19]"},{"why":"Defines the frequency-dependent beam splitter with double squeezed inputs that the paper compares against FC.","marker":"[20]"},{"why":"Proposes the EPR two-mode squeezing scheme that the paper compares against FC for heavy binaries.","marker":"[21]"},{"why":"Defines the KAGRA post-O5 high-frequency configuration parameters and the 85 m space constraint used throughout.","marker":"[35]"},{"why":"Identifies dephasing and locking precision as the main issue for short filter cavities, motivating the AFC/FDBS comparison and the paper's key caveat.","marker":"[37]"},{"why":"Gives the sideband-to-two-photon conversion and analytic filter-cavity bandwidth and detuning formulas that the optimization starts from.","marker":"[48]"},{"why":"Shows the analytic formulas fail when losses are significant, motivating the numerical particle-swarm optimization central to the claims.","marker":"[49]"}],"fun_headline_variants":["KAGRA's 85 m filter cavity boosts BNS detections by 23%","Optimized filter cavity beats EPR for KAGRA BNS searches","Filter cavity wins when quantum noise dominates KAGRA BNS","23% more BNS events with KAGRA's optimized filter cavity","Detuned cavity improves KAGRA BNS rate by 23%"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The quantitative comparison assumes an 85 m detuned filter cavity can be locked and operated without dephasing or locking errors that degrade the injected squeezing; this is not modeled in the FC calculation, even though the paper cites dephasing as the main problem for short filter cavities.","fun_headline_variants_meta":{"raw":{"variants":["KAGRA's 85 m filter cavity boosts BNS detections by 23%","Optimized filter cavity beats EPR for KAGRA BNS searches","Filter cavity wins when quantum noise dominates KAGRA BNS","23% more BNS events with KAGRA's optimized filter cavity","Detuned cavity improves KAGRA BNS rate by 23%"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001405,"raw_usage":{"total_tokens":5734,"prompt_tokens":1059,"completion_tokens":4675,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":675,"completion_tokens_details":{"reasoning_tokens":4577}},"tokens_in":675,"tokens_out":4675,"duration_ms":30640,"temperature":1.0,"reasoning_tokens":4577,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T12:00:26.656150+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Lock an 85 m detuned, over-coupled filter cavity with the proposed parameters and measure the squeezing degradation due to dephasing and locking noise at the operating point; if the degradation is large enough that less than about 15 dB of effective squeezing reaches the interferometer, the predicted at least 23% BNS detection-rate gain over FIS will not appear in the measured sensitivity.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the filter-cavity scheme for frequency-dependent squeezing that all FC comparisons build on."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows a single filter cavity suffices for RSE-configuration interferometers, justifying the 85 m single-cavity design."},{"cited_title":"Evans, L","cited_arxiv_id":null,"evidence_quote":"Supplies the empirical loss scaling and confocal-length normalization used to set realistic round-trip losses."},{"cited_title":"Corbitt, N","cited_arxiv_id":null,"evidence_quote":"Defines the amplitude filter cavity scheme that the paper compares against FC."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the frequency-dependent beam splitter with double squeezed inputs that the paper compares against FC."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Proposes the EPR two-mode squeezing scheme that the paper compares against FC for heavy binaries."},{"cited_title":"McCuller, S","cited_arxiv_id":null,"evidence_quote":"Identifies dephasing and locking precision as the main issue for short filter cavities, motivating the AFC/FDBS comparison and the paper's key caveat."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the sideband-to-two-photon conversion and analytic filter-cavity bandwidth and detuning formulas that the optimization starts from."},{"cited_title":"Whittle, K","cited_arxiv_id":null,"evidence_quote":"Shows the analytic formulas fail when losses are significant, motivating the numerical particle-swarm optimization central to the claims."}],"review_version":1}