{"id":"6e5b9f3c-f8c0-4192-aab1-4495aac4ec65","arxiv_id":"2607.05157","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"low","formal_verification":"none","parameter_count":3,"one_line_summary":"Multigrid MLMC yields up to O(10^5) variance reduction for connected correlators while torus probing with dilution substantially reduces cost for disconnected loops, confirming complementary regimes.","lead":"Two variance-reduction methods for expensive lattice-QCD traces are compared: multigrid multilevel Monte Carlo and a new torus-based probing scheme. Multigrid MLMC cuts cost for long-distance correlators; probing cuts cost for local disconnected loops, showing clear complementarity.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.5","headline":"Cost-reduction claim for multigrid MLMC rests on variance plots without full accounting of setup and multilevel sampling overhead.","rationale":"The reader's weakest assumption correctly flags limited ensemble and setup-parameter coverage, but the more immediate load-bearing gap is the missing end-to-end cost accounting that would convert the impressive variance collapse of G_{1,1} into a verified computational saving. The mathematical constructions (oblique projectors, torus coloring) appear sound and the reported variance numbers on this ensemble are credible; the paper itself already notes that multilevel gains for disconnected loops do not yet beat plain Hutchinson once overhead is considered. A single, fully costed run on the existing configuration would settle whether the connected-case claim survives the same scrutiny. Because that check is still missing, the CONDITIONAL verdict remains appropriate; no stronger rejection is warranted.","tokens_in":8959,"tokens_out":537,"duration_ms":4704,"concrete_test":"Recompute the connected correlator on the same IV ensemble using the optimal N_l of eq. (6) for a target variance ε^{2}, measure total linear solves (including the N_tv=28 setup solves) and wall-clock time for both plain Hutchinson and multigrid MLMC, and report the ratio. If the measured cost ratio is less than ~3\times at large t, the 'clear cost reduction' claim weakens.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The strongest claim (abstract, §4.1) asserts that the multigrid multilevel decomposition yields a variance reduction of up to O(10^5) at large t and 'translates into a clear cost reduction at fixed accuracy' for the connected correlator. Figure 1 and the surrounding text show only per-component variances of the G_{i,j} terms under a fixed N=500; they do not report the actual sample allocation N_l from the MLMC formula (eq. 6), the wall-clock or flop cost of the multigrid setup (N_tv=28 test vectors), or the relative cost C_l of a coarse-level solve versus a fine-level solve. Without those numbers it is possible that residual coarse-level variance plus setup overhead erase most of the claimed net saving, especially once the hierarchy must be rebuilt for each new configuration. The disconnected-loop results already illustrate that moderate fine-level variance reduction can fail to produce net cost improvement; the same risk is not ruled out for the connected case.","agreement_with_reader":"partial"},"referee_report":{"model":"grok-4.5","summary":"The manuscript studies two complementary variance-reduction techniques for stochastic trace estimation in lattice QCD: (i) multigrid multilevel Monte Carlo (MLMC), which builds an unbiased multilevel estimator from the recursive oblique-projector splitting induced by a Galerkin multigrid hierarchy (eqs. 8–12), and (ii) stochastic probing based on a new torus coloring (eq. 15) that uses substantially fewer colors than hierarchical probing at the same distance (Table 1). On a single RQCD IV configuration, the multilevel decomposition of the connected pseudoscalar two-point function yields fine-level variance reductions up to O(10^5) at large time separations (Fig. 1), while for disconnected fermion loops multilevel gains remain moderate (Fig. 2) and probing combined with full dilution produces a clear reduction in total solves versus target variance (Fig. 3). The authors conclude that deflation-type schemes are most effective for long-distance observables and probing for localized ones.","tokens_in":9204,"tokens_out":1301,"duration_ms":27463,"significance":"If the reported complementarity holds more generally, the paper supplies a practical guide for matching variance-reduction tools to the structure of lattice QCD observables, which is of direct use for precision disconnected and connected correlator campaigns. Strengths include an algebraically unbiased multilevel identity that reuses the existing multigrid solver hierarchy, an explicit and tableable torus coloring with fewer colors than hierarchical probing, and an honest demonstration that multilevel MLMC does not automatically improve cost for disconnected loops. The work also confirms the large fine-level variance collapse for connected correlators reported in arXiv:2412.06347. These are concrete, implementable contributions rather than purely formal ones.","major_comments":[{"comment":"The abstract and §4.1 assert that the multilevel decomposition 'translates into a clear cost reduction at fixed accuracy' for the connected pseudoscalar correlator, and §5 quantifies 'cost reductions of about one order of magnitude.' Figure 1 and the surrounding text report only per-component variances of the G_{i,j} terms under a fixed sample size N=500. Unlike the probing study (Fig. 3), the paper does not report the MLMC sample allocation N_l from eq. (6), the relative costs C_l of fine versus coarse solves, the multigrid setup cost for N_tv=28 test vectors, or a total-solve (or wall-clock) curve versus target variance. Because residual variance is concentrated in the coarsest term G_{3,3}, a net saving is plausible, but it is not demonstrated in this manuscript; the disconnected-loop discussion already shows that moderate fine-level variance reduction need not yield net cost improvem","section":"Abstract, §4.1, Fig. 1, §5"},{"comment":"All numerical results (variance plots, cost curves, and the complementarity conclusion) are obtained on a single IV configuration (Table 2) with a fixed multigrid setup (three levels, N_tv=28). The weakest assumption underlying the fine-level variance collapse is that these transfer operators already capture enough low-mode content for the observed reductions to be representative. Without at least a second ensemble, a different volume, or a sensitivity check in N_tv, the strength of the claim that 'deflation schemes are most effective for observables dominated by long distance propagation, while probing is most effective for localized quantities' remains limited. A short additional data set or an explicit caveat that the complementarity is demonstrated on one ensemble would make the conclusion proportionate to the evidence.","section":"§4 setup, Table 2, §§4.1–4.2, §5"}],"minor_comments":[{"comment":"The torus coloring (eq. 15) is a useful practical contribution, but the paper only compares color counts to hierarchical probing (Table 1), not variance reduction or cost at fixed accuracy. The outlook correctly flags a controlled benchmark; a sentence in §3 or §4.2 stating that such a comparison is left for future work would prevent readers from over-reading Table 1 as a performance comparison.","section":"§3, Table 1, §5"},{"comment":"Variance estimates themselves are obtained with fixed N=500. A brief remark on the statistical uncertainty of the reported variances (especially the O(10^5) ratio at large t) would strengthen confidence in Fig. 1.","section":"§4, Fig. 1"},{"comment":"Typographical issues: 'combinantion' should be 'combination' (end of §3); several run-on words appear in the compiled text (e.g. near the start of §4.2). Axis labels and legends in Figs. 1–3 should be checked for readability in the final PoS layout.","section":"§3, §4.2, Figs. 1–3"},{"comment":"The relation between the multilevel splitting used here and the Wick-contraction decomposition of Ref. [1] could be stated more explicitly in §2 or §4.1 so that the precise novelty relative to that work is clear.","section":"§2, §4.1"}],"recommendation":"major_revision","confidential_remarks":"This is a Lattice 2025 proceedings-style manuscript. The science is solid and the complementarity message is useful; the main obstacle to acceptance as written is the unsupported cost-reduction language for multigrid MLMC on the connected correlator. If the venue is PoS, a minor_revision path with softened cost claims might be enough; if the authors intend a longer journal version (e.g. CPC or PRD), the missing cost accounting and multi-ensemble checks should be required. No integrity or citation concerns."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The one thing worth knowing is that this paper gives a clean, side-by-side demonstration that multigrid multilevel Monte Carlo and a new torus coloring attack different parts of the variance in lattice-QCD traces. For the connected pseudoscalar correlator the multilevel split (eqs. 9–12, 17–20) collapses the expensive fine-level variance by up to O(10^5) at large t (Fig. 1); for disconnected loops the same multilevel only helps moderately, while probing + full dilution cuts the solve count substantially (Fig. 3). That complementarity is the real takeaway and matches the physics of long-range versus local contributions.\n\nWhat is new is modest but useful. Multigrid MLMC for traces already existed (Frommer et al. 2022; Gruber et al. arXiv:2412.06347); the paper confirms and extends those numbers on the RQCD IV ensemble. The genuine increment is the torus coloring (eq. 15 and Table 1) that needs roughly four times fewer colors than hierarchical probing at the same distance. The algebra is clean: the recursive oblique projectors give an unbiased estimator with no free parameters, and the variance plots are consistent with the claimed reductions under fixed N=500.\n\nThe soft spot is exactly the one the stress-test flags, and it is real but not fatal. Figure 1 shows per-component variances, not the optimal sample allocation N_l from the MLMC formula (eq. 6), not the relative costs C_l of coarse versus fine solves, and not the setup cost of the 28 test vectors. Without those numbers the abstract’s “clear cost reduction at fixed accuracy” is only partially supported; the disconnected case already shows that moderate fine-level gains can be wiped out by overhead. Single-ensemble, fixed-N variance estimates also leave generalization open. These are ordinary limitations of a short conference paper, not hidden flaws.\n\nAnyone running production correlators or disconnected loops will get value from the figures and the coloring formula. The math and citations are solid; the work is honest about where each method wins. I would send it to peer review without hesitation—referees can demand the missing cost tables—and I would cite the complementarity result and the torus coloring myself.","headline":"Solid numerical methods paper that cleanly shows multigrid MLMC and torus probing are complementary on two standard QCD observables; the O(10^5) variance claim is real on the given ensemble, but net-cost accounting is incomplete.","tokens_in":9836,"tokens_out":561,"would_cite":true,"duration_ms":4668,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"Multigrid multilevel Monte Carlo and torus-based probing cut variance in lattice QCD traces by complementary mechanisms: one for long-distance correlators, one for local loops.","keywords":["lattice QCD","trace estimation","multilevel Monte Carlo","multigrid","stochastic probing","variance reduction","disconnected loops","pseudoscalar correlator"],"falsifier":"Repeat the connected-correlator measurement on a second ensemble with different volume or quark mass using the same multigrid hierarchy; if the fine-level variance reduction falls well below O(10^3–10^5), or if the torus coloring requires substantially more colors than hierarchical probing at equal distance, the claimed complementarity fails.","tokens_in":9856,"feed_emoji":"⚛️","tokens_out":924,"duration_ms":7060,"temperature":0.7,"pith_summary":"Computing traces involving the inverse of the Wilson-Dirac operator is a bottleneck in lattice QCD, because the plain stochastic Hutchinson estimator only improves as the square root of the number of samples. This paper shows that two existing variance-reduction ideas become highly effective when adapted to the structure of that operator, and that each idea matches a different class of observables. Multigrid multilevel Monte Carlo recursively peels off coarse-grid corrections already available from the multigrid solver; for the connected pseudoscalar two-point function the fine-level variance collapses by as much as five orders of magnitude at large time separations, producing a clear net cost saving. Stochastic probing with a new torus coloring, combined with dilution, instead annihilates short-range off-diagonal contributions; for disconnected fermion loops it yields a substantial reduction in the number of linear solves that improves steadily with the number of probing vectors. The practical message is that deflation-style methods should be used when the observable is dominated by long-distance modes, while probing should be used when the dominant fluctuations are local.","feed_headline":"Multigrid and probing cut lattice-QCD trace costs by 10–10^5","feed_subtitle":"Deflation wins for long-distance correlators; torus probing wins for local loops","key_machinery":"Recursive multigrid splitting of the inverse via the oblique projectors already present in the multigrid hierarchy (eqs. 8–11), together with a distance-d torus coloring (eq. 15) that produces far fewer colors than hierarchical probing at the same distance.","core_discovery":"The multigrid multilevel decomposition of the inverse yields a variance reduction of up to O(10^5) for the connected pseudoscalar correlator at large time separations and a clear cost reduction at fixed accuracy, while for disconnected loops the same multilevel scheme gives only moderate gains and probing plus dilution is the method that produces a substantial, scalable cost saving. The two techniques are therefore complementary rather than interchangeable.","pith_inferences":["The same multigrid splitting should apply, with only minor changes, to other connected Wick contractions that involve products of two or more inverses.","Once the torus-coloring coefficients are tabulated for standard lattice sizes they become a free, reusable library resource for any code that already uses probing.","If the moderate multilevel gain observed for disconnected loops can be amplified by adding a few exact low modes, the two methods may become synergistic rather than merely complementary."],"forward_implications":["Connected long-distance correlators can be estimated at fixed accuracy with far fewer expensive fine-grid solves by shifting residual variance onto cheap coarse levels.","Disconnected local loops become cheaper by systematically increasing the number of torus-probing vectors plus dilution rather than by multilevel deflation.","The two techniques can be combined so that long-range and short-range variance contributions are attacked simultaneously.","Any observable whose variance is known to be dominated by either low modes or by local off-diagonal entries can now be assigned the matching reduction strategy a priori."],"fun_headline_variants":["Multigrid multilevel MC cuts lattice-QCD correlator variance by O(10^5)","Torus probing beats hierarchical coloring for lattice-QCD disconnected loops","Multigrid wins long-distance correlators; probing wins local fermion loops","Complementary variance cuts for lattice-QCD traces: multigrid vs torus probing","Multilevel deflation gives 10^5 variance drop on connected pseudoscalar correlators"],"cache_read_input_tokens":128,"weakest_assumption_plain":"That transfer operators built from only 28 test vectors on one 64\times32^{3} ensemble already capture enough low-mode content for the fine-level variance collapse to hold more generally, and that the torus-coloring coefficients found by exhaustive search stay near-optimal on other lattices.","fun_headline_variants_meta":{"raw":{"variants":["Multigrid multilevel MC cuts lattice-QCD correlator variance by O(10^5)","Torus probing beats hierarchical coloring for lattice-QCD disconnected loops","Multigrid wins long-distance correlators; probing wins local fermion loops","Complementary variance cuts for lattice-QCD traces: multigrid vs torus probing","Multilevel deflation gives 10^5 variance drop on connected pseudoscalar correlators"]},"model":"grok-4.5","effort":"low","cost_usd":0.004706,"raw_usage":{"total_tokens":1341,"prompt_tokens":787,"num_sources_used":0,"completion_tokens":104,"cost_in_usd_ticks":47060000,"prompt_tokens_details":{"text_tokens":787,"audio_tokens":0,"image_tokens":0,"cached_tokens":128},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":450,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":787,"tokens_out":104,"duration_ms":4018,"temperature":1.0,"reasoning_tokens":450,"cache_read_input_tokens":128,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-11T07:53:58.124598+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Repeat the connected-correlator measurement on a second ensemble with different volume or quark mass using the same multigrid hierarchy; if the fine-level variance reduction falls well below O(10^3–10^5), or if the torus coloring requires substantially more colors than hierarchical probing at equal distance, the claimed complementarity fails.","supporting_citations":[],"review_version":1}