{"id":"062c642e-39c5-4d7b-b687-e14fd4110fd1","arxiv_id":"2507.07816","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A JAX/GPU implementation of a long-transient gravitational-wave search is about 60 times faster per template than ATrHough and suggests that 10^8 to 10^9 templates suffice for near-optimal sensitivity.","lead":"A team of gravitational-wave analysts implemented a GPU-accelerated search for long-lived signals from newborn neutron stars. It runs about 60 times faster per template than an earlier method and may need far fewer templates, making all-sky follow-up searches practical.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Wide-sky feasibility claim rests on sensitivity measured with true T0 and sky position; the reported template reduction does not include the sky/T0 search cost.","rationale":"The reader's weakest assumption identifies the same load-bearing issue: the sensitivity results are obtained with templates that know the true T0 and sky position of each injection, while the abstract claims a capability that requires searching over those parameters. My independent reading of Section 3 confirms that the template banks cover frequency, braking index, and spindown timescale ranges, but not the sky-position or T0 dimensions. The reported d90 distances are therefore best-case sensitivity estimates, and the claimed 1-2 order template reduction applies only to the reduced parameter space. I do not see a problem with the speedup measurement itself; the benchmark is explicit about CPU/GPU and batch sizes. The concern is specifically that the headline feasibility claim is broader than what the injection campaign demonstrates. This supports the reader's conditional verdict rather than overturning it, so no change to the verdict is needed. A concrete follow-up test that marginalizes over sky and T0 would settle whether the wide-sky claim can be retained or should be limited.","tokens_in":4759,"tokens_out":2893,"duration_ms":35274,"concrete_test":"Repeat the sensitivity campaign of Section 3 with a template bank that also covers sky position (e.g., a coarse grid of 10x10 sky locations) and T0 (e.g., offsets over the 24-hour data duration), keeping the same total N=10^7 templates and using a noise-only Psi threshold. Recompute the d90 values. If d90 drops substantially or N must be increased beyond 10^9 to recover the fixed-sky/T0 result, the wide-sky/no-pinpointing claim is not supported; if d90 is unchanged, the claim is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central feasibility claim has two legs: a 60x per-template speedup on GPUs and a 1-2 order-of-magnitude reduction in the number of templates needed for near-optimal sensitivity. The speedup leg is a timing benchmark and is credible as reported. The template-count leg is the weak point. In Section 3, the injection protocol fixes the two most important track-defining parameters: \"Each template used the T0, and sky position values of the injection being searched for.\" Templates far from the injected signal are then excluded from the detection criterion. The d90 values in Fig. 4 are therefore conditional on knowing T0 and sky position exactly, not on finding the signal over the sky. The abstract nevertheless claims that the method \"enables wide regions of the sky to be covered, eliminating the need for precise pinpointing of mergers or supernovae.\" A wide-sky search must template over sky position and T0. Those dimensions are absent from the N=10^4-10^7 template banks used to derive the claimed 1-2 order template reduction. If the full search requires multiplying the bank by N_sky x N_T0, the claimed reduction relative to ATrHough could be eroded or reversed, and the sensitivity measured under perfect prior knowledge may not be representative. This is a load-bearing gap because the paper's stated deployment goal is wide-sky searches following poorly localized binary-neutron-star merger alerts.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents a GPU-accelerated implementation, written in Python with JAX, of a search for long-duration transient gravitational waves (tCWs) from newborn neutron stars. The authors report a per-template computing time of roughly 15 microseconds on GPU hardware, about 60 times faster than the previously published ATrHough method, and claim that the required template density is one to two orders of magnitude lower than ATrHough while achieving comparable sensitivity. Sensitivity is assessed by injecting simulated signals at 19 distances, building template banks of sizes N = 10^4 to 10^7, and computing 90% detectability distances from sigmoidal efficiency fits. The authors conclude that the method will enable wide-sky searches without precise electromagnetic localization of the progenitor.","tokens_in":5016,"tokens_out":3015,"duration_ms":37896,"significance":"If the claims are fully substantiated, the work would represent a practical computational advance for tCW searches, which are currently limited by the cost of template evaluation. The timing benchmark on A100 GPUs is encouraging and the idea of porting the search to JAX is a reasonable route to speedups. However, the paper's headline feasibility claim — that wide sky coverage is enabled and precise localization is unnecessary — is not supported by the reported sensitivity study, because the templates in that study use the true start time and sky position of each injection. The template-reduction claim relative to ATrHough also lacks a same-campaign comparison. The strengths are the clear timing measurements and the concrete injection setup; the weaknesses are in the extrapolation from the idealized search to the stated deployment scenario.","major_comments":[{"comment":"The abstract claims that the method 'enables wide regions of the sky to be covered, eliminating the need for precise pinpointing of mergers or supernovae,' but the sensitivity measurement does not test this capability. The text states that 'Each template used the T0, and sky position values of the injection being searched for' and that 'templates far away from the injected signal were excluded.' Consequently, the d90 values in Figure 4 are conditional on exact knowledge of the two parameters that define the track's sky and time placement. A real wide-sky search must tile over sky position and T0, which adds a multiplicative factor to the template bank and introduces mismatch losses that are not quantified here. This is a load-bearing gap for the central deployment claim and must be addressed, either by including sky and T0 in the injection/template search or by providing an explicit estimate of the added cost and sensitivity degradation.","section":"Section 3, injection protocol and Figure 4"},{"comment":"The statement that 'the number of templates N required is 1 to 2 orders of magnitude lower than what required by the ATrHough method' is not substantiated by any comparative measurement in this manuscript. No ATrHough sensitivity or template-density baseline is recomputed on the same injection campaign, and the only reference to ATrHough's performance is the per-template timing in Section 3. The extrapolation from N=10^7 to N=10^8-10^9 as 'sufficient to achieve maximum d90 results' is also based on only four data points, with no defined criterion for saturation. Please provide a direct comparison of template counts needed to reach a given d90 in the same scenario, or clearly restate the claim as an estimate based on external literature.","section":"Section 3, template-count comparison with ATrHough"},{"comment":"The detection criterion is not sufficiently specified for reproducibility. The text says 'templates far away from the injected signal were excluded' and that an injection is detected if a 'close' template has critical ratio above the noise maximum, but neither 'far away' nor 'close' is given a quantitative definition. Without a mismatch metric or distance threshold, the efficiency curves in Figure 3 and 4 cannot be reconstructed by an independent implementation. Please define the proximity criterion in terms of a parameter-space metric or mismatch.","section":"Section 3, detection criterion"}],"minor_comments":[{"comment":"The phrase 'Thank to the usage of JAX' in Section 3 contains a typo ('Thank' should be 'Thanks'); throughout the paper there are missing spaces in expressions such as 'wherefgw,0', 'thelalpulsar Make-fakedata v5code', and 'ω' formatted inconsistently.","section":"Abstract and Section 1"},{"comment":"The caption of Figure 2 says the histograms show 'for loop times per template' but does not explain what the loop consists of, how many templates per batch, or whether the timing includes JIT compilation, memory transfer, or only the kernel execution; this makes the 15 microseconds figure hard to interpret.","section":"Figure 2 caption"},{"comment":"The text says the full theoretical ranges for n and τ were used in the template banks, but Section 1 lists the ranges as 3 to 7 for n and 3500 to 35000 s for τ, while the injections use fixed n=5 and τ=10000 s; it would be helpful to state explicitly that the banks cover those ranges and how many templates per dimension are used.","section":"Section 3, parameter ranges"},{"comment":"The conclusion that 'the method developed demonstrates sensitivities comparable to those of the previous ATrHough method' is not directly evidenced in the paper, since no ATrHough detection efficiency or d90 comparison is shown; this sentence should be qualified or backed by a reference to a comparison.","section":"Conclusions"},{"comment":"Reference [9] is a URL-only citation for JAX; please include a version or access date, and consider citing the JAX paper or documentation more formally.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is a short proceedings contribution, so some brevity is expected, but the missing sensitivity analysis over sky position and T0 is fundamental to the stated application and not just an omitted detail. The template-reduction claim also needs a proper baseline. These issues can be fixed in a revision, but the manuscript as it stands overstates the readiness of the method for wide-sky searches. The timing results are useful and should be preserved."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Best quick take: the 60x per-template speedup on GPUs is the solid, standalone result. The timing benchmark is simple, reproducible in principle, and not entangled with the sensitivity analysis. That alone is worth noting for anyone building long-transient searches.\n\nThe paper does two other things well. The template-density study is a sensible way to ask how many templates are needed for near-optimal sensitivity, and comparing cubic-lattice and uniform-random banks at fixed N is a useful check. Using both normalized power and number-count statistics, and reporting the 10–15% distance penalty for ν, is honest.\n\nThe soft spot is exactly where the stress-test note lands. Section 3 says each template used the T0 and sky position of the injection, and templates far from the injection were excluded. That means the measured d90 values are conditional on knowing those values, not on finding the signal over sky and start time. Yet the abstract claims the method 'enables wide regions of the sky to be covered, eliminating the need for precise pinpointing.' The injection campaign does not exercise that capability. A wide-sky search must multiply the bank by N_sky × N_T0, and the claimed 1–2 order reduction in template count does not account for that cost. The reduction could be eroded or reversed. This is a load-bearing gap, not a cosmetic one.\n\nTwo smaller issues. The N=10^8–10^9 sufficiency extrapolation is stated without running those banks; that is fine for a proceedings note, but the '1 to 2 orders of magnitude lower than ATrHough' claim would be stronger with an apples-to-apples ATrHough baseline on the same injections. No code release is mentioned, so the speedup number cannot currently be checked.\n\nNone of this makes the paper worthless. The per-template timing result stands, and the template-density scaling is a genuinely new quantitative observation. The paper just does not support its headline deployment claim as written.\n\nFor whom: GW data-analysis specialists working on long-transient or continuous-wave follow-up searches. A serious referee should see it, but with a request to either add the sky/T0 search cost to the template-count estimate or soften the abstract and conclusions accordingly. I'd engage with it in that spirit.","headline":"A credible per-template speedup and a soft template-count claim; the wide-sky feasibility statement is not backed by the injection protocol.","tokens_in":5535,"tokens_out":2050,"would_cite":false,"duration_ms":23964,"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 JAX/GPU implementation of long-transient gravitational-wave searches runs about 60 times faster per template than the previous method, and needs 1–2 orders of magnitude fewer templates to reach the same sensitivity.","keywords":["gravitational waves","long-transient signals","newborn neutron stars","GPU acceleration","JAX","template banks","Hough transform","detectability distance"],"falsifier":"Repeat the injection study without giving templates the true sky position and $T_0$, searching over those parameters instead, and compare the 90% detectability distance at $10^7$ templates with the ideal-track curve; if it drops steeply, the wide-sky-without-pointing conclusion is falsified.","tokens_in":4565,"feed_emoji":"🌌","tokens_out":10243,"duration_ms":100964,"temperature":0.7,"pith_summary":"The paper presents a GPU-accelerated pipeline for searching long-duration gravitational-wave transients emitted by newborn neutron stars. It reports a per-template speedup of roughly 60 compared with the previous ATrHough method, dropping from about 1 ms to 15 microseconds on GPUs, and finds that template banks of $10^8$ to $10^9$ entries are enough to reach near-optimal sensitivity. That reduces the cost of a wide-sky search from months to days, making follow-up observations after a merger alert practical without precise localization. If correct, the approach would let the same detection capability be deployed with much smaller computing budgets.","feed_headline":"GPU search runs 60 times faster on newborn-neutron-star signals","feed_subtitle":"That speedup, plus needing 100x fewer templates, makes wide-sky transient searches practical.","key_machinery":"The load-bearing object is the template, a model of a signal's time-frequency track governed by the msMagnetarWaveform frequency law $f_{\\mathrm{gw}}(t)=f_{\\mathrm{gw},0}\\left(\\frac{t-T_0}{\\tau}+1\\right)^{1/(1-n)}$. Each template is scored against short-Fourier-transform data by a normalized-power or number-count statistic summarized as a critical ratio $\\Psi$, and the JAX implementation compiles the scoring into batched GPU operations. The paper's central measurement is how the 90% detectability distance grows with template-bank size for cubic-lattice and uniform-random banks, which lets it infer the minimum bank density for near-optimal sensitivity.","core_discovery":"The paper reports that a JAX/GPU search pipeline evaluates a long-transient gravitational-wave template in about 15 microseconds on an A100 GPU and about 0.3 ms on a CPU, a 60-fold improvement over the roughly 1 ms per template of ATrHough. In simulated injections over distances from 0.1 to 3.1 Mpc, the 90% detectability distance with $10^7$ templates approaches the maximum attainable when templates exactly follow the signal tracks, and the trend indicates that $10^8$–$10^9$ templates would saturate that sensitivity. Since this is 1–2 orders of magnitude fewer templates than ATrHough requires, the authors conclude that wide-sky coverage is achievable without precisely knowing the merger or supernova location.","pith_inferences":["The reported sensitivity study supplies the true sky position and coalescence time to every template and excludes distant templates, so the conclusion that wide-sky searches need no precise localization is an extrapolation rather than a directly measured result.","A decisive follow-up test would rerun the injections with sky position and $T_0$ left free; the observed drop in detectability distance would quantify how much sensitivity the unsearched degrees of freedom actually cost.","The 15-microsecond figure was measured on a specific GPU with large batch sizes; the practical speedup on smaller or consumer-grade GPUs could differ by a factor of several.","If the sigmoidal trend in detectability distance flattens before $10^8$ templates, the inferred saturation point would move upward, but the qualitative advantage over ATrHough would remain."],"forward_implications":["A search that would have required about two months of computing with the previous method can be completed in roughly one day on GPUs.","Template-bank sizes of about $10^8$ to $10^9$ appear sufficient to reach near-maximum sensitivity in the studied parameter range, one to two orders of magnitude fewer than ATrHough needed.","The normalized-power statistic gives 90% detectability distances 10–15% longer than the number-count statistic, so the choice of statistic directly affects search reach.","The speedup holds on both CPUs and GPUs, with per-template times of 0.3 ms and 15 microseconds respectively, so the method remains practical when GPU resources are limited.","The same JAX/GPU batching strategy can be transferred to other long-duration gravitational-wave signals with different frequency evolution models."],"supporting_citations":[{"why":"It supplies the JAX library, whose GPU computation and just-in-time compilation enable the per-template speedup reported in the paper.","marker":"[9]"},{"why":"It defines the ATrHough method that serves as the baseline for timing and template-bank-density comparisons.","marker":"[14]"},{"why":"It provides the msMagnetarWaveform model that sets the frequency evolution used by the templates and injections.","marker":"[11]"},{"why":"It supplies the code used to generate the simulated signals in the sensitivity injections.","marker":"[10]"},{"why":"It gives the Advanced LIGO design sensitivity curve used to color the Gaussian noise into which signals were injected.","marker":"[7]"}],"fun_headline_variants":["GPU speeds up newborn-neutron-star wave search 60x","Search for newborn neutron star waves 60x faster on GPUs","Wide-sky gravitational wave search made practical with GPUs","GPU search needs 100x fewer templates for neutron star waves","Fast GPU method scans sky for newborn neutron star signals"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The sensitivity results come from templates that were given the true coalescence time and sky position of each injected signal, with distant templates excluded, so the claimed ability to search the whole sky without precise pointing is not directly demonstrated by the measurements.","fun_headline_variants_meta":{"raw":{"variants":["GPU speeds up newborn-neutron-star wave search 60x","Search for newborn neutron star waves 60x faster on GPUs","Wide-sky gravitational wave search made practical with GPUs","GPU search needs 100x fewer templates for neutron star waves","Fast GPU method scans sky for newborn neutron star signals"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000313,"raw_usage":{"total_tokens":1755,"prompt_tokens":897,"completion_tokens":858,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":513,"completion_tokens_details":{"reasoning_tokens":772}},"tokens_in":513,"tokens_out":858,"duration_ms":8670,"temperature":1.0,"reasoning_tokens":772,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T18:31:37.972521+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Repeat the injection study without giving templates the true sky position and $T_0$, searching over those parameters instead, and compare the 90% detectability distance at $10^7$ templates with the ideal-track curve; if it drops steeply, the wide-sky-without-pointing conclusion is falsified.","supporting_citations":[{"cited_title":"2018, http://github.com/google/jax","cited_arxiv_id":null,"evidence_quote":"It supplies the JAX library, whose GPU computation and just-in-time compilation enable the per-template speedup reported in the paper."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It defines the ATrHough method that serves as the baseline for timing and template-bank-density comparisons."},{"cited_title":"2017, Tech","cited_arxiv_id":null,"evidence_quote":"It provides the msMagnetarWaveform model that sets the frequency evolution used by the templates and injections."},{"cited_title":"2018, Tech","cited_arxiv_id":null,"evidence_quote":"It gives the Advanced LIGO design sensitivity curve used to color the Gaussian noise into which signals were injected."}],"review_version":1}