{"id":"7b421927-ed86-4bb3-90b4-2a458af756c3","arxiv_id":"1909.00756","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":3,"one_line_summary":"Z_V for HISQ vector currents is exact and condensate-free in RI-SMOM and in a modified ratio scheme, but the standard RI'-MOM scheme has O(1%) condensate contamination.","lead":"This lattice QCD paper shows that the RI-SMOM renormalisation scheme gives clean, condensate-free renormalisation factors for vector currents, while the popular RI'-MOM scheme carries an O(1%) nonperturbative error that can be removed by a simple ratio modification. It also provides the first RI-SMOM test of quenched QED effects on the vector current renormalisation factor, finding effects below 0.1% for the HISQ action.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified; the exact Ward-Takahashi identity, not the form-factor benchmark, carries the SMOM exactness claim, and the benchmark caveat does not change the verdict.","rationale":"The paper's main claim is a theorem-like consequence of the exact vector Ward-Takahashi identity, not an empirical pattern that could be invalidated by the form-factor benchmark. Equation (8) is exact configuration-by-configuration, the SMOM projector is chosen so that the conserved current gets Z_V=1, and Eq. (9) is the standard operator identity relating a nonconserved lattice current to the conserved current up to discretisation effects. The numerical checks in Figs. 2 and 3 support the implementation, and the difference plots provide a useful but secondary validation. The reader's weakest-assumption, namely reliance on Ref. [11]'s F(0) benchmark, is a genuine caveat: on the finest lattices the F(0) values appear to have been inferred from a perturbative fit rather than measured directly, so the comparison there is not fully independent. However, even a bias in those inferred values would not overturn the Ward-identity argument; it would only weaken one piece of supporting evidence. The same conclusion would be reached by an independent charge-conservation check of the SMOM-renormalised current. I therefore see no load-bearing concern that requires changing the ACCEPT verdict, although a direct superfine F(0) measurement would strengthen the empirical case. The agreement_with_reader is partial because the benchmark caveat is real but, in contrast to the reader's framing, it is not the load-bearing support for the central claim.","tokens_in":30213,"tokens_out":16113,"duration_ms":195225,"concrete_test":"Compute the zero-momentum vector form factor F(0) for the local HISQ current directly on the superfine set-7 ensemble, using the same methodology as Ref. [11] without any perturbative extrapolation, and refit the difference in Eq. (29). If the leading condensate coefficient remains within roughly 0.05 of zero and the continuum-limit value of Z_loc^V(SMOM)-Z_loc^V(F(0)) shifts by less than about 0.001, the inferred-benchmark caveat is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim that RI-SMOM Z_V is free of condensate contamination is established by the exact lattice vector Ward-Takahashi identity (Eq. 8) together with the standard operator relation Eq. (9), which says the renormalised local current coincides with the conserved current up to O(a^2). Because the SMOM procedure is constructed to respect the Ward identity, Z_cons^V(SMOM)=1 is expected and verified to 0.05% in Fig. 3, and the same identity protects the local-current factor from condensate terms that survive the continuum limit. The agreement with the form-factor benchmark is a consistency check rather than the load-bearing step. The one caveat, also noted by the reader, is that the Section IV D comparison on the finest ensembles uses Z_loc^V(F(0)) values from Ref. [11] that were inferred from a perturbative-plus-discretisation fit, not directly measured on those ensembles. If that fitted benchmark had an unrecognised condensate contribution, the difference plots in Figs. 6, 9 and 15 would be shifted. This is a real limitation of the numerical demonstration, but it is not load-bearing for the main conclusion, which follows from the Ward-Takahashi identity independently.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the renormalisation of flavour-diagonal vector currents in lattice QCD using the HISQ action on MILC n_f=2+1+1 ensembles, with emphasis on momentum-subtraction schemes. It derives the exact lattice vector Ward-Takahashi identity for the HISQ conserved current and verifies it on a single gauge configuration. It then shows that the RI-SMOM scheme gives Z_cons^V = 1 to better than 0.05% and that Z_loc^V from RI-SMOM differs from the form-factor value Z_loc^V(F(0)) only by discretisation effects, with the leading condensate coefficient constrained to -0.020(44). In contrast, the standard RI'-MOM scheme yields Z_cons^V and Z_loc^V with O(1%) condensate contamination, which can be removed by using ratios of conserved-to-local vertex functions in the RI'-MOMRc and SMOM-gamma_mu,Rc schemes. A first study of quenched QED effects in the RI-SMOM scheme shows the QED correction to Z_loc^V is below 0.1%.","tokens_in":30505,"tokens_out":9246,"duration_ms":95683,"significance":"If the conclusions hold, RI-SMOM provides an exact, fully nonperturbative Z_V that requires no momentum window and no condensate fit, which is practically valuable for precision calculations such as the hadronic vacuum polarisation contribution to the muon anomalous magnetic moment. The load-bearing argument is the exact Ward-Takahashi identity, not the external benchmark: the paper verifies Eq. (8) to double precision on a single configuration, confirms Z_cons^V(SMOM)=1 to 0.05%, and supports the local-current result with a controlled difference fit. The internal consistency across schemes and lattice spacings, and the explicit quantitative diagnosis of RI'-MOM contamination, are strengths. The main caveat is that parts of the comparison to the form-factor benchmark use values inferred from a fit in Ref. [11], but this does not undermine the Ward-Takahashi-based reasoning.","major_comments":[],"minor_comments":[{"comment":"The comparison to Z_loc^V(F(0)) on fine and superfine ensembles uses benchmark values inferred from a fit in Ref. [11] rather than values measured directly on those ensembles. The text should state explicitly how the uncertainties of those benchmark values are propagated into Figs. 6, 9, and 15, and acknowledge that a common contamination in the benchmark would shift the difference plots. This does not affect the Ward-Takahashi-based argument for the SMOM result, but it is a limitation of the numerical demonstration.","section":"Section IV D"},{"comment":"The name 'Ward-Takashashi' appears in the Introduction, Section II heading, and Section IV A; it should be 'Ward-Takahashi'.","section":"Sections I and II"},{"comment":"The tree-level factor used for the conserved current in the RI-SMOM-gamma_mu scheme is only described verbally as a modification of Eq. (24) to the SMOM kinematics; writing the explicit analogue of Eq. (24) would make the scheme reproducible without guesswork.","section":"Section IV F"},{"comment":"The lower panel of Fig. 6 labels the horizontal axis as 'a^2 [fm]', which is dimensionally inconsistent; it should be 'a^2 [fm^2]'.","section":"Section IV D / Fig. 6"},{"comment":"The sentence 'This is not true configuration by configuration' in the discussion of Z_cons^V(SMOM) is slightly confusing because Eq. (8) is exact on every configuration; the statement should clarify that the ensemble average is needed for the replacement in Eq. (14), not for the Ward-Takahashi identity itself.","section":"Section III"}],"recommendation":"minor_revision","confidential_remarks":"The paper is a careful methodological study with a sound central argument; the only substantive caveat is the reliance on an inferred form-factor benchmark for part of the numerical comparison, and this is not load-bearing. I am happy for the report to be sent to the authors as is."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is a solid lattice methods paper, and the central claim holds up: for the HISQ local vector current, the RI-SMOM renormalisation factor is free of nonperturbative condensate contamination. What carries the argument is not the comparison to the form-factor benchmark but the exact lattice Ward-Takahashi identity, which they verify on a single configuration to double precision and in two gauges. The numerical demonstration that Z_cons^V(SMOM)=1 to better than 0.05% is convincing, and the difference plots against Z_V(F(0)) are fit well by discretisation terms alone, with the leading condensate coefficient constrained to zero. This is a genuinely new quantitative demonstration for the HISQ action, and the proposed RI'-MOM Rc ratio scheme is a simple, effective fix for a scheme that otherwise has O(1%) contamination. The quenched-QED study is exploratory but careful, with sanity checks against perturbation theory and finite-volume effects.\n\nThe soft spots are minor. The finest-lattice comparisons in Figs. 6, 9, and 15 use Z_V(F(0)) values from Ref. [11] that were inferred from a perturbative-plus-discretisation fit rather than measured directly on those ensembles. If that benchmark carried an unrecognised condensate contamination, the difference plots would shift. But this is not load-bearing: the exactness claim follows from the Ward-Takahashi identity alone. The RI'-MOM condensate fit has a fair number of nuisance parameters, and the leading condensate coefficient is only about three sigma from zero on its own, but the need for condensate terms is unambiguous (chi2/dof jumps from 0.6 to 7.7 without them), so the qualitative conclusion is secure. The section on the axial current is a side note, but it is honest about the small residual difference in the SMOM scheme.\n\nThis paper deserves a serious referee. It is a careful, well-documented methods study with high-precision numerical evidence, and it will be directly useful to anyone renormalising nonconserved currents in lattice QCD, especially for sub-percent HVP calculations. I would cite it and bring it to a reading group focused on renormalisation or on lattice methods for g-2.\n\nRecommendation: accept after minor revisions, mainly to soften the claims where the form-factor benchmark is doing some work and to make the fit priors and the provenance of the finest-lattice benchmark values a bit more transparent.","headline":"A careful and convincing lattice-QCD methods paper: the RI-SMOM exactness claim is carried by the Ward-Takahashi identity, the numerics are excellent, and the soft spots are minor.","tokens_in":31098,"tokens_out":2066,"would_cite":true,"duration_ms":24096,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81T25","81T80","81V05"],"pacs":["11.15.Ha","12.38.Gc"],"model":"deepseek-v4-flash","headline":"The renormalisation factor for the nonconserved local HISQ vector current obtained in the RI-SMOM scheme is exact, free of nonperturbative condensate contamination, and differs from the charge-conservation determination only by…","keywords":["lattice QCD","vector current renormalisation","RI-SMOM scheme","RI-prime-MOM scheme","Ward-Takahashi identity","HISQ action","condensate contamination","quenched QED"],"falsifier":"Examine the difference $Z_{\\rm loc}^V({\\rm SMOM})-Z_{\\rm loc}^V(F(0))$ on a finer lattice or with a different quark action and check whether it extrapolates to exactly zero in the continuum limit, since a nonzero intercept would falsify the claim. Equivalently, measure the coefficient of the $1/\\mu^2$ term in $Z_{\\rm loc}^V({\\rm SMOM})$ with higher statistics; the paper constrains it to $-0.020(44)$, so a value several standard deviations from zero would show contamination that the Ward-Takahashi protection was supposed to remove.","tokens_in":30017,"feed_emoji":"⚛️","tokens_out":8178,"duration_ms":67857,"temperature":0.7,"pith_summary":"This paper establishes that the renormalisation factor for the nonconserved local HISQ vector current obtained in the RI-SMOM momentum-subtraction scheme is exact, in the sense that it carries no nonperturbative condensate contamination and therefore agrees with the charge-conservation form-factor determination up to cleanly removable discretisation effects. The same protection is shown to hold for the one-link point-split vector current. By contrast, the standard RI-prime-MOM scheme is shown to carry about one percent condensate contamination in both conserved and local currents, and a modified ratio version, called RI-prime-MOM Rc, is proposed and verified to be safe. The paper also provides a first nonperturbative study of quenched QED corrections to the vector current renormalisation in the RI-SMOM scheme, finding an effect below 0.1% for the HISQ action.","feed_headline":"RI-SMOM scheme gives condensate-free vector current renormalisation","feed_subtitle":"These factors match charge conservation up to discretisation effects; standard RI-prime-MOM carries about 1% contamination.","key_machinery":"The load-bearing mechanism is the exact lattice vector Ward-Takahashi identity, which in momentum space reads $-2i a^{-1}\\sin(aq_\\mu/2)\\Lambda_\\mu^+ = -S^{-1}(p_1)+S^{-1}(p_2)$ and holds configuration by configuration in any gauge. For the HISQ action the conserved current is a one-link-plus-three-link Naik-improved operator, and the RI-SMOM scheme, a momentum-subtraction scheme with symmetric kinematics $p_1^2=p_2^2=q^2=\\mu^2$ and a projector constructed from this identity, therefore forces $Z_V=1$ for the conserved current. Any nonconserved current renormalised by a ratio to the conserved vertex is then protected from the condensate contaminations that would otherwise appear as $1/\\mu^2$ terms; this Ward-Takahashi-protected ratio is the central object, and for the RI-prime-MOM scheme the same ratio, called RI-prime-MOM Rc, is what restores safety.","core_discovery":"Using the exact lattice vector Ward-Takahashi identity for the HISQ action, the authors show that any renormalisation procedure built on that identity must assign $Z_V=1$ to the conserved current; they verify this explicitly for RI-SMOM kinematics, with $Z_{\\rm cons}^V({\\rm SMOM})$ equal to 1 to better than 0.05% at every scale. Because the renormalised local current coincides with the conserved current up to $​O(a^2)$ discretisation effects, the RI-SMOM value $Z_{\\rm loc}^V({\\rm SMOM})$ inherits the same protection: its difference from the form-factor value $Z_{\\rm loc}^V(F(0))$ extrapolates to zero in the continuum limit, and the fit constrains any $1/\\mu^2$ condensate term to $-0.020(44)$, statistically consistent with zero. The standard RI-prime-MOM scheme does not use the identity, and its $Z_V$ for both the conserved and local currents shows a clear condensate contribution with leading coefficient $0.154(54)$, so that naive use brings roughly one percent systematic errors; defining $Z_V$ instead as the ratio of local to conserved vertex functions, the Rc variant, removes this contamination.","pith_inferences":["The paper's result that there is no lower limit on $\\mu$ suggests a step-scaling strategy in which $Z_V$ is computed at low $\\mu$, where discretisation effects are smallest, and then run to higher scales; the authors demonstrate a two-$\\mu$ combination but do not pursue full step scaling.","The ratio-to-conserved-current construction generalises: any lattice current with an exactly conserved counterpart could be renormalised in any momentum-subtraction scheme by normalising to the conserved vertex, potentially extending the same protection to other operators with an exact Ward-Takahashi identity.","For actions whose pure-QCD $Z_V$ is far from 1, the quenched-QED shift in $Z_V$ should scale roughly as $(1-Z_V)\\,\\alpha_{\\rm QED}$; if confirmed, this would let collaborations estimate QED corrections to renormalisation factors without a dedicated calculation.","A direct test of the Ward-Takahashi protection would be to compute the difference $Z_V({\\rm SMOM})-Z_V(F(0))$ at $\\mu$ values below 1 GeV on large volumes, where the paper's data show smaller discretisation effects and any hidden condensate term would first become visible."],"forward_implications":["$Z_{\\rm loc}^V({\\rm SMOM})$ can be used at any momentum scale $\\mu$ without fitting away condensate effects, so there is no need for a restricted window of $\\mu$ values; the only trade-off is between statistical error and discretisation error.","Standard RI-prime-MOM determinations of $Z_V$ carry about one percent systematic error unless condensate terms are included in a multi-$\\mu$, multi-$a$ fit, whereas the RI-prime-MOM Rc ratio removes that error at the cost of computing the conserved-current vertex.","For staggered quarks, the local axial current renormalisation $Z_A$ can be set equal to the local vector result $Z_V$, avoiding a separate momentum-subtraction calculation that would reintroduce chiral-symmetry-breaking condensate contamination.","Quenched QED corrections to $Z_V$ in the RI-SMOM scheme are below 0.1% for the HISQ action and are describable by the known perturbative coefficient, so QED effects can be included nonperturbatively in precision matrix-element calculations.","The conclusions are not specific to the HISQ action: any improved action with an exact lattice Ward-Takahashi identity should enjoy the same protection in RI-SMOM and in the Rc-type ratio schemes."],"supporting_citations":[{"why":"Supplies the form-factor benchmark $Z_{\\rm loc}^V(F(0))$ that the momentum-subtraction values are compared against.","marker":"[11]"},{"why":"Defines the RI-SMOM scheme as the momentum-subtraction construction based on the Ward-Takahashi identity, giving $Z_V=1$ for the conserved current.","marker":"[15]"},{"why":"Introduces the general nonperturbative renormalisation method and the original RI-prime-MOM scheme whose standard $Z_V$ is shown to carry condensate contamination.","marker":"[12]"},{"why":"Provides the continuum loop conversion factors from RI-prime-MOM to MS that are needed to interpret the standard RI-prime-MOM results.","marker":"[13]"},{"why":"Defines the HISQ action and its Naik-improved derivative, from which the one-link-plus-three-link conserved current is constructed.","marker":"[4]"},{"why":"Establishes the staggered-quark momentum-subtraction techniques, valence-mass extrapolation, and ensemble framework used for the numerical determinations.","marker":"[2]"},{"why":"Supplies the staggered-quark spin-taste implementation and tree-level vertex factors for the point-split and conserved currents.","marker":"[21]"}],"fun_headline_variants":["RI-MOM's 1% bias fixed by ratio method","RI-SMOM gives condensate-free Z_V; RI-MOM has 1% bias","Ratio method removes condensate bias from RI-MOM Z_V","Ward identity protects RI-SMOM from condensate pollution"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes that the form-factor benchmark used for comparison, fixed by charge conservation, is itself exactly free of nonperturbative contamination; any hidden error in that benchmark would shift the difference plots and be inherited by the claim that the momentum-subtraction results are clean.","fun_headline_variants_meta":{"raw":{"variants":["RI-MOM's 1% bias fixed by ratio method","RI-SMOM gives condensate-free Z_V; RI-MOM has 1% bias","Ratio method removes condensate bias from RI-MOM Z_V","Ward identity protects RI-SMOM from condensate pollution"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001619,"raw_usage":{"total_tokens":6535,"prompt_tokens":1126,"completion_tokens":5409,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":742,"completion_tokens_details":{"reasoning_tokens":5331}},"tokens_in":742,"tokens_out":5409,"duration_ms":305133,"temperature":1.0,"reasoning_tokens":5331,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T05:36:47.200820+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Examine the difference $Z_{\\rm loc}^V({\\rm SMOM})-Z_{\\rm loc}^V(F(0))$ on a finer lattice or with a different quark action and check whether it extrapolates to exactly zero in the continuum limit, since a nonzero intercept would falsify the claim. Equivalently, measure the coefficient of the $1/\\mu^2$ term in $Z_{\\rm loc}^V({\\rm SMOM})$ with higher statistics; the paper constrains it to $-0.020(44)$, so a value several standard deviations from zero would show contamination that the Ward-Takahashi protection was supposed to remove.","supporting_citations":[{"cited_title":"Wilson’s Operator Expansion: Can It Fail?","cited_arxiv_id":null,"evidence_quote":"Supplies the form-factor benchmark $Z_{\\rm loc}^V(F(0))$ that the momentum-subtraction values are compared against."},{"cited_title":"Renormalization and running of quark mass and field in the regulariza- tion invariant and MS-bar schemes at three loops and four loops,","cited_arxiv_id":null,"evidence_quote":"Defines the RI-SMOM scheme as the momentum-subtraction construction based on the Ward-Takahashi identity, giving $Z_V=1$ for the conserved current."},{"cited_title":"Three Topics in Renormalization and Improvement","cited_arxiv_id":"1103.1323","evidence_quote":"Introduces the general nonperturbative renormalisation method and the original RI-prime-MOM scheme whose standard $Z_V$ is shown to carry condensate contamination."},{"cited_title":"Column 3 gives results using the RI-SMOM scheme","cited_arxiv_id":null,"evidence_quote":"Establishes the staggered-quark momentum-subtraction techniques, valence-mass extrapolation, and ensemble framework used for the numerical determinations."},{"cited_title":"Lattice Fermions: Species Doubling, Chiral Invariance, and the Triangle Anomaly,","cited_arxiv_id":null,"evidence_quote":"Supplies the staggered-quark spin-taste implementation and tree-level vertex factors for the point-split and conserved currents."}],"review_version":1}