{"id":"3b6155e3-c827-4cab-8110-9e6e4e82f234","arxiv_id":"1908.09771","paper_version":3,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"A dynamical lattice QCD calculation using Ioffe-time pseudo-distributions produces an isovector nucleon valence PDF and first two moments at M_pi ~ 400 MeV, with estimates of lattice-spacing and finite-volume systematics.","lead":"This lattice QCD paper computes the nucleon's parton distribution function, the map of quark momentum inside a proton, using Ioffe-time pseudo-distributions on three ensembles at a pion mass near 400 MeV. The calculation is the first unquenched one of this kind and estimates both lattice-spacing and finite-volume systematic errors.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Higher-twist contamination of the largest-z data is in tension with the all-z simultaneous PDF fit in §4.2.","rationale":"The reader's weakest assumption is precisely the leading-twist dominance that justifies fitting all z2 separations simultaneously. My stress-test found that the paper internally flags the largest-z data as potentially higher-twist contaminated in Section 4.1, while Section 4.2 then uses those same data in the global PDF fit. This is the most load-bearing weak point because the highest moments and the large-x behavior of the PDF are controlled by the large-ν data, which only exist at large z. The paper is otherwise admirably candid: it does not claim a continuum-extrapolated PDF, it discusses the insufficiency of two lattice spacings, and it estimates finite-volume effects. Those limitations are appropriately reflected in a conditional verdict. Since my concern is essentially the same one identified by the reader, and it does not move the verdict away from CONDITIONAL, the appropriate recommendation is UNCHANGED. The proposed test directly targets whether the higher-twist contamination is numerically significant using only data already present in the paper.","tokens_in":31735,"tokens_out":6792,"duration_ms":79642,"concrete_test":"Reanalyze the published bare matrix elements in Tables 2-6 by fitting the reduced pseudo-ITD to M(ν,z2) = M_LT(ν) + z2 C(ν), where M_LT is the leading-twist form used in Section 4.2 and C(ν) is a low-order polynomial, using the jackknife covariance. If C(ν) is nonzero at more than 2σ, or if repeating the moment extraction with z/a ≥ 5 excluded shifts the matched second moment by more than the jackknife error, then the all-z simultaneous fit is not protected against higher-twist contamination.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central extraction of the valence PDF and its moments relies on leading-twist dominance of the reduced pseudo-ITD after the double ratio and NLO matching, allowing all separations up to z/a = 8 to be fit simultaneously in Section 4.2. The paper's own Section 4.1, however, states that only the largest-z2 points have enough Ioffe-time range to constrain the higher terms in the Taylor expansion, and that 'since these data potentially have significant higher twist corrections, these results should be considered questionable.' The small-z points are described by c1 and c2 alone. Nevertheless, Section 4.2 includes the questionable large-z data in the simultaneous PDF fit, even though those are precisely the data that constrain the large-ν (large-x) behavior. The observed z2 independence of the reduced pseudo-ITD is empirical support, but it is not a proof: the double-ratio denominator can cancel higher-twist contributions that are correlated between numerator and denominator while leaving residual effects in the data used for the PDF. If those residual effects are present, the extracted x dependence and the higher moments are biased, which would weaken the paper's claim to have identified the systematics of the PDF extraction.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript presents a lattice QCD calculation of the unpolarized nucleon valence PDF using the Ioffe-time pseudo-distribution (pseudo-ITD) formalism. The authors use three 2+1-flavor Wilson-clover ensembles with lattice spacings a = 0.127 fm and a = 0.094 fm and two volumes, all at a pion mass of roughly 400 MeV. Matrix elements are extracted with a Feynman-Hellmann method and momentum smearing, and the reduced pseudo-ITD is formed through a double ratio that cancels the Wilson-line renormalization constants. The paper extracts the first two PDF moments from a Taylor/Vandermonde analysis, fits the Ioffe-time distribution to a four-parameter PDF ansatz, and studies discretization and finite-volume effects by comparing the two lattice spacings and two volumes. The central claims are that this is the first dynamical-fermion pseudo-PDF calculation, that the continuum and finite-volume systematics can be estimated with these ensembles, and that the resulting x-dependence and moments can be compared with global fits.","tokens_in":32068,"tokens_out":7601,"duration_ms":77940,"significance":"If the analysis is robust, this paper is a valuable step for the pseudo-PDF program: it demonstrates the method with dynamical fermions, uses momentum smearing and Feynman-Hellmann extraction to control excited states, and addresses two sources of systematic uncertainty (lattice spacing and volume) that were not both considered in earlier pseudo-PDF work. The paper is transparent: bare matrix elements are tabulated for all ensembles, jackknife errors are propagated, and the authors explicitly refrain from claiming a continuum-extrapolated PDF. The renormalization-group-invariant double ratio and the NLO matching kernel provide a clean route from the lattice matrix element to the MS PDF. The main risks are the reliance on leading-twist dominance at large separations and the dependence of the extracted x-dependence on a single model ansatz, both of which are acknowledged in the text but not fully quantified.","major_comments":[{"comment":"The simultaneous fit of all z/a separations in §4.2 is in tension with the statement in §4.1 that the higher Taylor coefficients, which control the large-ν behavior, are only constrained by the largest-z data, and that \"Since these data potentially have significant higher twist corrections, these results should be considered questionable.\" The large-ν points in the PDF fit of §4.2 are precisely the large-z points, so the extracted x-dependence and higher moments are directly sensitive to the same higher-twist contamination that the paper flags as questionable. The observed z² independence of the reduced pseudo-ITD is empirical support, but the double ratio only partially cancels O(z² Λ²_QCD) effects and is not a proof. I request a systematic stability test: repeat the PDF fit excluding points with z/a > 5 and z/a > 4, and/or include an explicit higher-twist term such as M(ν,z²) = M_LT(ν) + z² H(ν), and report whether the valence PDF and moments change within errors.","section":"§4.1–§4.2"},{"comment":"The central result—the x-dependence of the valence PDF—is never displayed: Figs. 17–19 show only the fitted Ioffe-time distribution Q(ν) compared with the lattice data, and the fitted parameters a, b, c, d of Eq. (4.4) are not reported for any ensemble. As a result, the reader cannot quantitatively assess the extracted x-space PDF or its uncertainty, nor can the moments be reproduced from the fit. Please include a plot of x f_v(x) (or f_v(x)) with a jackknife error band for each ensemble, and give a table of the fitted parameters with their correlations.","section":"§4.2, Figs. 17–19"},{"comment":"The x-dependence is obtained from a single four-parameter ansatz, Eq. (4.4), taken from CJ/MSTW global fits. The paper acknowledges that the choice of parameterization introduces model-dependent bias, but it does not quantify that bias. Since the x-dependence is one of the paper's main claims, the analysis should include at least one alternative functional form (e.g., P(x)=1+c x+d x², or a form without the √x term in Eq. (4.4)) and propagate the difference between the resulting PDFs as a systematic uncertainty. The quoted χ²/d.o.f. values (2.0–2.5) are not small enough to argue that the data alone select this particular form.","section":"§4.2, Eq. (4.4)"}],"minor_comments":[{"comment":"The abstract states that \"continuum limit and infinite volume extrapolation systematic errors of the PDF are considered,\" but the body explicitly declines to produce a continuum-extrapolated PDF and studies these effects on the reduced pseudo-ITD rather than on the final PDF or moments; please rephrase to avoid overstating the scope.","section":"Abstract and §4.3–4.4"},{"comment":"The statement that the double ratio is exactly unity at ν=0 \"with no possible higher twist effects and lattice spacing errors\" is true by construction, but it may be clearer to note explicitly that this is a normalization condition that fixes the zeroth moment rather than a physical cancellation.","section":"§3.5, text near Eq. (3.21)"},{"comment":"The polynomial interpolation uses degree six (real) and degree five (imaginary) for data reaching ν ≈ 9.4; the paper says a cubic spline was checked, but does not report the numerical comparison. A sentence quantifying the difference between the two interpolations would help the reader judge the systematic error introduced here.","section":"§3.7, Eq. (3.24)"},{"comment":"The residual z² dependence of the matched MS moments is said to be negligible, but no numerical χ² or p-value for a constant fit is given; a quantitative statement would strengthen the justification for combining all z² in the PDF fit.","section":"§4.1, Fig. 16"},{"comment":"The finite-volume ansatz M_Inf = M_Latt + C_L exp(-m(L-z)) is presented as an assumption, and the conclusions note that the functional form still needs to be determined. In Fig. 21 the finite-volume difference is shown, but not divided by the combined statistical error; please add a quantitative estimate of the significance of the volume effect.","section":"§4.4, Eq. (4.9)"}],"recommendation":"major_revision","confidential_remarks":"The paper is a solid technical advance and is likely to be influential in the lattice-PDF community. The major comments concern the robustness and presentation of the central x-dependence result: the all-z fit in §4.2 should be tested against a z-cut or a higher-twist model, the x-space PDF should be shown, and the model dependence of the ansatz should be quantified. These are localized, addressable changes that do not require new ensembles. I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know: this is the first unquenched pseudo-PDF calculation that tries to estimate both continuum and finite-volume systematics, and it is a genuinely careful piece of work. The authors get a reduced pseudo-ITD at M_pi ~ 400 MeV on three ensembles, extract the first two moments, and compare to global fits. They never claim a continuum-extrapolated PDF, and they are explicit that two lattice spacings are not enough. That restraint is real and earns credit.\n\nWhat's new: the dynamical ensembles, the momentum-smearing + Feynman-Hellmann extraction for a nonlocal operator, and the first multi-ensemble benchmark for the pseudo-PDF approach. The analysis itself is also above the usual bar: the double ratio for renormalization cancellation is clean, the excited-state fits use a motivated form and are shown, and the bare matrix element tables are included. The NLO matching kernel is cited rather than derived, which is appropriate.\n\nThe soft spots are real but mostly smaller than they first look. The stress-test concern about leading-twist dominance is partly correct: Section 4.1 explicitly says the large-z points that constrain higher moments \"potentially have significant higher twist corrections\" and should be considered questionable, yet Section 4.2 fits all z simultaneously. The empirical z^2 independence of the reduced ITD is supporting evidence, and the double ratio cancels some higher-twist effects, but the tension deserves a robustness test: drop the largest z and see if the PDF fit moves. That should be a referee request, not a rejection. The single-PDF-ansatz issue is also a real limitation; the paper acknowledges it but does not test an alternative parameterization. No code or data release is mentioned, which makes reproducibility harder, though the tables help. The unblinded choice of fit ranges to minimize chi2/dof is a mild concern, not a fatal one, since the main qualitative conclusions do not hinge on a few chi2 points.\n\nWho this is for: lattice QCD practitioners working on PDFs, and phenomenologists who want a concrete lattice benchmark at heavier pion mass. It deserves a serious referee. I would send it to review with requests for a large-z cut study and, if possible, a second functional form for the PDF.","headline":"First dynamical, two-spacing/two-volume pseudo-PDF calculation; honest and useful even though the x-shape rests on a single ansatz and on leading-twist assumptions that the paper itself flags.","tokens_in":32620,"tokens_out":1439,"would_cite":true,"duration_ms":19036,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["12.38.Gc","11.15.Ha","14.20.Dh"],"model":"deepseek-v4-flash","headline":"The first dynamical-fermion Ioffe-time pseudo-distribution calculation yields a nucleon valence PDF and its first two moments, with estimates of continuum and finite-volume systematics.","keywords":["parton distribution functions","lattice QCD","Ioffe time","pseudo-distributions","nucleon structure","operator product expansion","continuum extrapolation","finite-volume effects"],"falsifier":"Split the data by separation: fit the PDF and the moments using only the points with $z/a\\le 4$ and only those with $z/a\\ge 5$, after applying the same evolution and matching, and check whether the two results agree within statistical errors. A disagreement beyond errors would show that the separation dependence is not fully removed by the double ratio and that the simultaneous all-$z$ fit is contaminated.","tokens_in":31534,"feed_emoji":"⚛️","tokens_out":18548,"duration_ms":163908,"temperature":0.7,"pith_summary":"The paper aims to show that the Ioffe-time pseudo-distribution method, previously tested in the quenched approximation, works with dynamical fermions and can produce a nucleon valence PDF together with an estimate of the main systematic errors. Using three ensembles with two lattice spacings and two volumes at a pion mass around 400 MeV, the authors form a double ratio that cancels the ultraviolet renormalization of the non-local operator, evolve the data at next-to-leading order, and extract both the $x$-dependence of the valence quark distribution and the first two of its moments. The two lattice spacings give an estimate of the continuum extrapolation error, and the two volumes give an estimate of finite-volume effects. The moments are found to be higher than the phenomenological values from an NLO global fit, consistent with the heavier-than-physical pion mass. If this is right, the pseudo-PDF approach is a viable path to PDFs from first principles with a controlled error budget.","feed_headline":"First dynamical-lattice nucleon PDF from Ioffe-time data","feed_subtitle":"Two lattice spacings and two volumes set the error budget; valence-PDF moments are compared with global fits.","key_machinery":"The central object is the reduced pseudo-ITD\n$$\n\\overline{\\mathcal M}(\\nu,$z^{2}$)=\\frac{\\mathcal $M^{0}$(\\nu,$z^{2}$)/\\mathcal $M^{0}$(\\nu,0)}{\\mathcal $M^{0}$(0,$z^{2}$)/\\mathcal $M^{0}$(0,0)},\n$$\nwhere $\\mathcal M^0(\\nu,z^2)$ is the bare forward nucleon matrix element of $\\bar\\psi(z)\\gamma_4 W(z;0)\\psi(0)$ with $\\nu=p\\cdot z$. The double ratio cancels the Wilson-line renormalization constants and the quark-number normalization, and the result is renormalization-group invariant. The $z^2$ dependence is governed at one loop by the Altarelli–Parisi kernel $B(u)$, and the NLO matching to the $\\overline{\\rm MS}$ Ioffe-time distribution uses the kernel $K(u,z^2\\mu^2,\\alpha_s)$. The $x$-space PDF is recovered by fitting the matched real-part ITD to a valence-quark parameterization, while the moments $b_n(z^2)$ of the pseudo-PDF are matched to the PDF moments $a_n(\\mu^2)$ through the multiplicative Wilson coefficients $K_n$.","core_discovery":"The paper claims that the reduced Ioffe-time pseudo-distribution—a double ratio of forward matrix elements of a non-local quark bilinear with a straight Wilson line—carries the same dominant short-distance information as the light-cone Ioffe-time distribution, with the ultraviolet and power divergences canceled by the ratio. On three dynamical ensembles the authors find that this reduced function is nearly independent of the separation $z^2$ at fixed Ioffe time $\\nu$, so all data with $z/a$ up to 8 can be evolved with the one-loop kernel and matched at next-to-leading order to the $\\overline{\\rm MS}$ Ioffe-time distribution. They then fit the matched real part with a valence-quark parameterization to obtain the $x$-dependent PDF, and independently obtain the first two PDF moments—the average momentum fraction and the second moment—from the Taylor expansion of the same function using operator product expansion Wilson coefficients. The moments are systematically higher than those of the NLO global fit benchmark at this pion mass, as expected for $m_\\pi\\simeq 400$ MeV, and are consistent with one NNLO set within its errors.","pith_inferences":["Editorial inference: the near-$z^2$ independence suggests that the double-ratio subtraction absorbs most higher-twist contamination at the tested separations; if that survives at finer lattices, large-$z$ data, the only points sensitive to higher moments, can be used without per-separation higher-twist fits.","Editorial inference: the gap between the lattice moments and the NLO global-fit values at $m_\\pi\\simeq 400$ MeV gives a quantitative anchor for the pion-mass dependence; repeating the calculation at lower pion masses would show whether the moments approach the physical values linearly in $m_\\pi^2$ or with a different rate.","Editorial inference: the same double-ratio plus NLO-matching pipeline could be transferred to the pion PDF or to sea-quark and gluon distributions, using the imaginary component of the pseudo-ITD that isolates the $q+\\bar q$ combination; a pion benchmark would give an independent test of the leading-twist assumption.","Editorial inference: because the Feynman–Hellmann extraction reaches its plateau at earlier Euclidean times, this setup may reduce the cost of physical-pion-mass simulations where excited-state contamination is the dominant bottleneck."],"forward_implications":["If the central claim holds, future pseudo-PDF computations can report a continuum and finite-volume error budget from two lattice spacings and two volumes, rather than quoting only statistical errors.","The near $z^2$-independence of the reduced pseudo-ITD justifies fitting all separations simultaneously with a single PDF Ansatz, turning an ill-posed Fourier inversion into a stable fit.","The first two PDF moments can be computed from the same correlation functions without an inverse Fourier transform, providing a direct low-cost check against global fits.","The combination of momentum smearing and Feynman–Hellmann extraction extends the usable Ioffe-time range to $\\nu\\simeq 9$, which is what makes the higher moments and the large-$x$ region accessible at all."],"supporting_citations":[{"why":"Supplies the Lorentz-invariant pseudo-PDF and pseudo-ITD formalism, including the canonical x support used throughout.","marker":"[32]"},{"why":"First quenched lattice implementation; defines the reduced pseudo-ITD double ratio and its evolution, which this paper extends to dynamical fermions.","marker":"[34]"},{"why":"Derives the OPE relation between pseudo-ITD moments and PDF moments through Wilson coefficients, used for the first two moments.","marker":"[47]"},{"why":"Provides the one-loop z^2 evolution equation applied to the reduced pseudo-ITD data.","marker":"[59]"},{"why":"Provides the NLO matching kernel connecting the reduced pseudo-ITD to the MS-scheme ITD.","marker":"[60]"},{"why":"Introduces the momentum smearing technique used to improve high-momentum nucleon signals.","marker":"[50]"},{"why":"Establishes the Feynman–Hellmann matrix-element extraction method that reduces excited-state contamination.","marker":"[66]"},{"why":"NLO global fit used as the phenomenological benchmark for the extracted PDF moments.","marker":"[69]"},{"why":"Analyzes finite-volume effects for spatially nonlocal operators, motivating the two-volume comparison.","marker":"[74]"},{"why":"Supplies the 2+1-flavor clover ensembles at two lattice spacings and two volumes used for all numerical results.","marker":"[64]"}],"fun_headline_variants":["Dynamical lattice nucleon PDF from Ioffe-time data","Nucleon PDF and moments from Ioffe-time pseudo-distributions","Two-spacing, two-volume nucleon PDF from Ioffe-time data","Ioffe-time pseudo-distributions give nucleon PDF and moments"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The calculation assumes that after the double ratio cancels the ultraviolet divergences, the remaining dependence on the quark-antiquark separation is small enough to be described by the one-loop perturbative evolution, so that data at all separations up to eight lattice spacings can be fit together; if non-perturbative short-distance effects at the largest separations are large enough to produce a systematic drift, the extracted PDF and the higher moments would be biased.","fun_headline_variants_meta":{"raw":{"variants":["Dynamical lattice nucleon PDF from Ioffe-time data","Nucleon PDF and moments from Ioffe-time pseudo-distributions","Two-spacing, two-volume nucleon PDF from Ioffe-time data","Ioffe-time pseudo-distributions give nucleon PDF and moments"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.003053,"raw_usage":{"total_tokens":11533,"prompt_tokens":881,"completion_tokens":10652,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":497,"completion_tokens_details":{"reasoning_tokens":10572}},"tokens_in":497,"tokens_out":10652,"duration_ms":69369,"temperature":1.0,"reasoning_tokens":10572,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:03:33.785649+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Split the data by separation: fit the PDF and the moments using only the points with $z/a\\le 4$ and only those with $z/a\\ge 5$, after applying the same evolution and matching, and check whether the two results agree within statistical errors. A disagreement beyond errors would show that the separation dependence is not fully removed by the double ratio and that the simultaneous all-$z$ fit is contaminated.","supporting_citations":[{"cited_title":"Parton distribution functions from reduced Ioffe-time distributions","cited_arxiv_id":"1801.03023","evidence_quote":"Provides the NLO matching kernel connecting the reduced pseudo-ITD to the MS-scheme ITD."},{"cited_title":"Finite-volume effects due to spatially non-local operators","cited_arxiv_id":"1805.01034","evidence_quote":"Analyzes finite-volume effects for spatially nonlocal operators, motivating the two-volume comparison."},{"cited_title":"Edwards, B","cited_arxiv_id":null,"evidence_quote":"Supplies the 2+1-flavor clover ensembles at two lattice spacings and two volumes used for all numerical results."}],"review_version":1}