PDF at small $x$ in the non-perturbative region

M. L. Nekrasov

REVIEW 4 major objections 5 minor 17 references

PDF at small $x$ in the non-perturbative region

T0 review · 4 major / 5 minor · reviewed 2026-08-03 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper derives a power-law small-x parton density from a binary splitting cascade, with the exponent fixed by the splitting probability, and shows that adding parton fusion saturates the density at a value set by the fusion probability.

desk verdict Tidy branching-cascade toy for the small-x power law, but the saturation result is assumed, not derived. read the letter →

arxiv 2601.12489 v3 pith:MAHL3MLR submitted 2026-01-18 hep-ph hep-th

M. L. Nekrasov This is my paper · ORCID

classification hep-phhep-th

keywords partondistributionfunctionsmall-xphysicsmodelbranchingcascadefusionsaturationnon-perturbativeQCDgenerationladder

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper tries to show that the small-x parton density of a fast-moving proton can be understood without perturbation theory, just by counting partons in a branching cascade. In the cascade, each splitting divides longitudinal momentum in half and occurs with probability w, so the average count in generation n is (2w)^n; translating generation number into momentum fraction gives x f(x) = (1-w)x^{-δ_w} with δ_w = ln(2w)/ln2. Adding fusion of same-generation partons with probability v makes the recurrence nonlinear and forces the density per unit rapidity to saturate at (1-w+v)^2/(4v) once x reaches about v^{1/δ_w}. If the derivation is sound, the small-x exponent and the saturation plateau are tied to two model parameters, and saturation arises from a critical density rather than from a virtuality scale. A sympathetic reader would care because this offers a non-perturbative, mechanism-level explanation of saturation that can be compared with data.

What carries the argument

The binary generation ladder: partons are indexed by generation n, with longitudinal fraction x_n=2^{-n}, and the conversion identity Δx_n=x_n Δn turns discrete occupation numbers into the continuous relation x f(x)=N_n. The fusion term in the balance equation and the parabolic expression for the number remaining per generation identify the maximum-density point and the plateau height.

What would settle it

Measure the gluon density x f(x) over a range of small x where fusion is negligible; the model requires x f(x/2)/(x f(x)) = 2w to be constant for every halving in x. If this ratio drifts with x instead of staying flat, the fixed-exponent power law fails; likewise, if the per-rapidity density dN/dy keeps growing with decreasing x past the predicted saturation point instead of slowing to a plateau, the fusion-saturation mechanism is falsified.

Watch

Extended reading notes

Core claim

In the author's model, partons are organized by generations: the parent proton parton is generation 0, and every splitting produces two daughters in the next generation, each carrying exactly half the parent's longitudinal momentum. With splitting probability w, the average number of partons that appear in generation n is (2w)^n, and after the 1-w fraction that do not split is removed, the average number remaining is (1-w)(2w)^n. Because x_n=2^{-n} and the interval Δx_n equals x_n for one generation step, the continuous PDF satisfies x f(x_n)=(1-w)(2w)^n, hence x f(x)=(1-w)x^{-δ_w}, δ_w=ln(2w)/ln2, over the moderate-small-x region. When fusion of two partons in the same generation is include

Load-bearing premise

The derivation assumes every splitting divides the parent's longitudinal momentum exactly in half, so that all partons of generation n share x=2^{-n} and the conversion Δx_n=x_n Δn holds; if real splittings are asymmetric or momentum fractions spread within a generation, the power exponent δ_w and the saturation scale change.

Editorial extensions

If this is right

With w>1/2, the total parton multiplicity grows as (P/μ)^{δ_w}, a power law in momentum; for w=1/2 it reduces to logarithmic growth ~ln(P/μ), marking the transition to branching cascades.
In the moderate-x regime, x f(x) is a pure power law, so f(x)~x^{-1-δ_w}; the exponent is tied to the splitting probability and lies in the range -1 < -δ_w < 0.
When fusion is included, saturation appears at x_s∼v^{1/δ_w}, with saturated density per unit rapidity (1-w+v)^2/(4v), both controlled by the fusion probability v.
The initial proton momentum needed for slow partons to form a saturated medium is estimated as P_s/μ∼v^{-1/δ_w}.
The model reproduces, at a qualitative level, the saturation phenomenon predicted by previous perturbative QCD analyses, but explains it as a density effect rather than a virtuality effect.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The author does not connect the splitting probability w to a measured quantity; a testable extension would identify w with the effective intercept extracted from deep-inelastic data and then predict the high-energy multiplicity exponent δ_w=ln(2w)/ln2, which could be checked in proton-proton and electron-proton data.
Because the model treats partons as longitudinally identical and ignores transverse structure, it predicts the same saturation height for protons of different transverse size; future measurements of the small-x gluon density in protons versus larger nuclei could test this density-only mechanism against a transverse-area mechanism.
The discrete identity x f(x_{n+1})/x f(x_n)=2w gives a direct experimental falsifier: measuring the gluon density ratio at x and x/2 across a decade should be a constant; any x-dependence would force the model to allow asymmetric splitting.
If fusion is stronger than assumed, the number remaining per generation rises and then falls to zero; the paper leaves open whether the system then stabilizes into a saturated state or collapses, an explicit model extension not settled here.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Colleague,

The pure-splitting part of this paper is a tidy exercise: with the explicit equal-halving assumption, the average occupation number (2w)^n gives x f(x) ~ x^{-δ_w}, δ_w = ln(2w)/ln 2, and momentum conservation checks out exactly. That part is internally consistent and clearly presented. The author also deserves credit for stating plainly that w is a free parameter and that fixing it from pomeron intercepts is illustrative, not predictive.

The soft spot is the saturation claim, and it is load-bearing. Section 4 does not solve the nonlinear recurrence (7). It drops ΔN to get the parabola (18), locates the maximum at \tilde N = (1-w+v)/(2v), and identifies the corresponding n_s by inserting the no-fusion form \tilde N_n ≈ (2w)^n. The problem is that at that n_s, \tilde N is O(1/v), and the fusion term in (7) is also O(1/v) — same parametric order as the splitting term. So the no-fusion approximation is not valid up to n_s, and the saturation scale is effectively assumed, not computed. After the maximum, the text admits 'our model provides no information about the behavior of the system' and then imposes the plateau as 'the simplest option.' Calling that 'saturation is detected' overstates what has been shown. The stress-test note is right on this.

There is also a circularity in the power-law exponent: δ_w is defined via w, and w is then chosen to reproduce δ_w = 0.1 or 0.3 from pomeron intercepts. So the slope is an input, not an output. The equal-halving assumption is strong; asymmetric splittings would change the x-dependence. The paper acknowledges some of this but not the circularity.

Where that leaves the paper: as a toy model of the branching cascade, it is a clean and honest demonstration, though the branching multiplicity itself goes back to Polyakov. As a derivation of saturation, it fails. The qualitative discussion of why saturation need not be tied to Q^2 is worth thinking about. But the central advertised result is unsupported.

I would bring this to a reading group only as an example of overclaiming from a toy model. I would not cite it in my own work. For peer review: I would desk reject it in its current form. If the author replaces the assumed plateau with a controlled treatment of (7) — even a numerical solution — then it would deserve another look.

Referee Report

4 major / 5 minor

Summary. The manuscript presents a two-parameter parton model for the small-x PDF in the non-perturbative region. Partons are organized into generations; each splitting is assumed to divide the parent's longitudinal momentum exactly in half, so generation n has x_n = 2^{-n}. With a constant splitting probability w and no fusion, the mean occupation of generation n is N_n=(1-w)(2w)^n, leading to x f(x)=(1-w)x^{-\delta_w} with \delta_w=\ln(2w)/\ln 2. Parton fusion with probability v is then added through the nonlinear recurrence Eq. (7). The paper derives an upper bound from probability conservation, analyzes a truncated balance equation, and proposes that the density saturates at n_s with x_s~v^{1/\delta_w} and dN/dy~(1-w+v)^2/(4v), after which complete stabilization is assumed. The final section compares these results qualitatively with BFKL, GLR, and BK expectations.

Significance. If the saturation derivation were valid, the paper would offer a simple, non-perturbative argument that small-x saturation is a parton-density phenomenon rather than a virtuality-scale phenomenon. The pure-splitting part is internally consistent: momentum conservation is explicitly checked in Eq. (1), and the conversion from discrete generations to a continuous PDF is transparent. The paper is also candid in places: it states that Eq. (7) is not solved and that the model provides no information beyond n_s. However, the central advertised result is not actually derived. The power-law exponent is essentially a reparametrization of the assumed binary-splitting rule, the saturation scale is obtained by inserting the no-fusion solution into a relation where fusion is not negligible, and the plateau is imposed by hand. These are load-bearing defects, so the paper cannot be recommended for publication in its present form.

major comments (4)

[§4, Eqs. (19)–(20)] The saturation scale is not derived. Equation (19) defines n_s by eN_{n_s}=(1-w+v)/(2v), but eN_n is the solution of Eq. (7), which the paper states is unknown. The text then says “Assuming that eN_{ns}→(2w)^{ns} as v→0,” i.e. it substitutes the no-fusion form into (19). At the resulting n_s one has (2w)^{n_s}∼1/v. In Eq. (7) the fusion source is at least v eN_{n+1}(eN_{n+1}-1)/2 ∼ 1/(2v), the same parametric order as the splitting term 2w eN_{n-1}. Thus the no-fusion approximation is not uniformly valid up to n_s, and x_s∼v^{1/δ_w} is an assumption, not a consequence of the recurrence.
[§4, after Eq. (18); Abstract] The saturation plateau is imposed, not derived. Eq. (18) gives N_n as a parabola in eN_n; once eN_n exceeds eN_{ns}, N_n decreases to zero. The text calls this decreasing option “formally consistent” but unsatisfactory, then selects “the simplest option is complete stabilization,” and explicitly admits “our model provides no information about the behavior of the system” beyond n_s. Therefore the constant dN/dy≈(1-w+v)^2/(4v) in Eq. (22) is an external stabilization hypothesis. The Abstract’s statement that “the phenomenon of saturation of the parton density is detected” overstates what the model establishes.
[§2, Eq. (6)] The power-law exponent is built into the discrete kinematics. Since x_n=2^{-n} and N_n=(1-w)(2w)^n by construction, substituting n=ln(1/x)/ln 2 gives x f(x)=(1-w)x^{-δ_w} with δ_w=ln(2w)/ln 2. The numerical comparison in §2 then fixes w by assuming δ_w=0.1 or 0.3 from the soft/BFKL Pomeron. Thus the model does not independently predict the small-x slope; it re-expresses an assumed input. This does not invalidate the pure-splitting kinematics, but it substantially reduces the force of the first advertised result.
[§3–§4, Eqs. (17)–(21)] The approximations leading to Eq. (21) are not controlled. Eq. (17) discards ΔN with the justification that ΔN=O(v^2). For occupation numbers near the would-be maximum, eN_n=O(1/v), and the discarded terms are v^2 C_{eN_n-2}+⋯=O(1), not O(v^2) uniformly. While this is subleading relative to N_{ns}=O(1/v), it means that Eq. (21) is obtained without a quantitative error estimate. More importantly, the preceding substitution of the no-fusion form for eN_n into (19) is the step that actually determines n_s, and that substitution is unjustified, as noted above.

minor comments (5)

[Abstract] “PDF at small xin” and similar spacing issues in the abstract and body should be corrected.
[§4] “in the letter case” should be “in the latter case.”
[§5] “parton distribution fusion” appears to be a typo for “parton distribution function.”
[§3, Eqs. (8)–(9)] The notation C_N is ambiguous: it is defined as N(N-1)/2, but the compact sum uses C_{N-2k}. Spell out that C_m is the binomial coefficient m(m-1)/2 or write inom{m}{2} for clarity.
[§4] The mixed notation eN_{ns} and n_s should be clarified; it is easy to misread eN_{ns} as eN_n times s.

Circularity Check

1 steps flagged · score 6.0 of 10

Saturation plateau is an ansatz, not a model prediction; the advertised saturation 'detection' reduces to the complete-stabilization assumption.

other [Section 4, after Eq. (21); summarized in Abstract]
"Unfortunately, in the letter case our model provides no information about the behavior of the system. However, we can take into account that when approaching n_s from below the growth of N_n slows down significantly before stopping completely. Given this, at least a slowdown in growth can be expected at n > n_s. The simplest option is complete stabilization, when N_n becomes a constant approximately equal to N_ns. In terms of continuous functions this would mean that xf(x) with a very small x would take a constant value of x_s f(x_s)=N_ns."

The abstract claims 'the phenomenon of saturation of the parton density is detected and a model estimate of its value in this regime is obtained,' but the plateau is not derived from the model recurrence (7). The paper explicitly states that beyond n_s the model provides no information, then inserts 'the simplest option is complete stabilization.' Thus dN/dy=(1-w+v)^2/(4v) is the assumed stabilization value, not a prediction; the advertised saturation result is equivalent to the added ansatz by construction.

full rationale

The free-evolution result x f(x)=(1-w)x^{-\delta_w} with \delta_w=\ln(2w)/\ln 2 is a direct consequence of the binary-ladder definitions x_n=2^{-n} and \tilde N_n=(2w)^n, so it is a model consequence rather than a circular reduction; the authors leave w free and calibrate it to pomeron intercepts for illustration. There is no load-bearing self-citation or imported uniqueness. The significant issue is the saturation claim: it rests on the explicit 'simplest option' of complete stabilization after the model disclaims knowledge beyond n_s, making the plateau value an input ansatz. Additionally, Eq. (20) locates n_s by substituting the no-fusion form (2w)^{n_s} into the maximum condition even though the paper has just stated the exact solution of Eq. (7) is unknown and fusion contributions are not small in that regime; the scale x_s~v^{1/\delta_w} is therefore an assumed crossover rather than a derived one. These are limitations affecting the asserted prediction, with the plateau being the clearest reduction of an output to an input assumption.

Assumptions & free parameters 3 free parameters · 7 assumptions · 0 invented entities

All central outputs are functions of the two free probabilities w and v plus the discrete half-splitting ladder. The saturation plateau is an added postulate, not a consequence of the recurrence. No new physical entities are introduced; the model is a re-parametrization of branching-process inputs.

free parameters (3)

w (splitting probability) = 0.54 or 0.62 in examples
Free model parameter; examples are obtained by setting δ_w=0.1 or 0.3 from pomeron intercepts. The central exponent δ_w=ln(2w)/ln2 is determined by w.
v (fusion probability) = not numerically set
Free parameter controlling saturation density ∝1/v and saturation momentum x_s∼v^{1/δ_w}; no model or data determination.
μ (minimum parton momentum) = ≈300 MeV in estimates
Lower cutoff where partons merge with vacuum fluctuations; sets maximum generation n̄ and x_min; chosen by hand.

assumptions (7)

domain assumption Partons in a fast-moving proton are quasi-free and form a static distribution in the transverse plane (Feynman parton model).
§1; foundational premise of the model.
ad hoc to paper Each splitting divides momentum exactly in half; all partons of generation n have x_n=2^{-n}.
§2; this is the weakest assumption—it fixes the x↔n map used to derive the power-law exponent.
ad hoc to paper Partons split independently with a common absolute probability w, requiring w≥1/2 for branching.
§2 after Eq. (1); defines the branching-process counting N_n=(2w)^n.
domain assumption Partons cannot have longitudinal momentum below μ; below this they merge with vacuum fluctuations, so cascades terminate at n̄.
§1-2; provides the cutoff and the estimate n̄≈13 for P=3 TeV, μ=300 MeV.
ad hoc to paper Two partons in the same generation fuse with probability v, and fusions across generations are negligible.
§3; basis of the nonlinear recurrence (7).
ad hoc to paper After reaching n_s, the parton density stabilizes at N_n≈N_ns.
§4; the 'saturation' claim is an assumption, explicitly acknowledged as 'the simplest option is complete stabilization'.
domain assumption Only one parton type (gluon-like) is considered and all partons descend from a single parent parton.
§1; simplifying limitations acknowledged by the author.

how reviews work

0 comments

Cite this review

Pith. "Pith review of PDF at small $x$ in the non-perturbative region." pith.science (2026). https://pith.science/paper/MAHL3MLR

@misc{pith2026260112489,
  author       = {Pith},
  title        = {Pith review of: PDF at small $x$ in the non-perturbative region},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MAHL3MLR}},
  note         = {Machine review of arXiv:2601.12489}
}

abstract

Parton distribution function (PDF) at small $x$ in a fast-moving proton is investigated within an upgraded parton model that includes parton splitting with branching cascades and parton fusion. In the region of moderately small $x$, we obtain a power-law behavior of the parton density $x f(x)$ with an exponent proportional to the logarithm of the probability of parton splitting. Taking into account parton fusion leads to a nonlinear equation for the PDF. In the region of very small $x$, the phenomenon of saturation of the parton density is detected and a model estimate of its value in this regime is obtained. The results are compared with those obtained previously based on the analysis of equations in logarithmic approximations of perturbative QCD.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

17 extracted references · 1 linked inside Pith

[1]

Feynman, Photon-hadron Interactions, W.A

R.P. Feynman, Photon-hadron Interactions, W.A. Benjamin 1972

1972
[2]

Gribov, L.N

V.N. Gribov, L.N. Lipatov, Deep inelastice −pscattering in perturbation theory, Sov. J. Nucl. Phys. 15 (1972) 438

1972
[3]

Dokshitzer, Calculation of the Structure Functions for Deep Inelastic Scat- tering ande +e− Annihilation by Perturbation Theory in Quantum Chromody- namics, Sov

Yu.L. Dokshitzer, Calculation of the Structure Functions for Deep Inelastic Scat- tering ande +e− Annihilation by Perturbation Theory in Quantum Chromody- namics, Sov. Phys. JETP 46 (1977) 641

1977
[4]

Altarelli, G

G. Altarelli, G. Parisi, Asymptotic Freedom in Parton Language, Nucl. Phys. B 126 (1977) 298

1977
[5]

L. V. Gribov, E. M. Levin, M. G. Ryskin, Semihard Processes in QCD, Phys. Rept. 100 (1983) 1

1983
[6]

Mueller, J.Qiu, Gluon Recombination and Shadowing at Small Values ofx, Nucl

A.H. Mueller, J.Qiu, Gluon Recombination and Shadowing at Small Values ofx, Nucl. Phys. B268 (1986) 427. 11

1986
[7]

B. L. Ioffe, V. A. Khoze and L. N. Lipatov, Hard Processes: Phenomenology, Quark-Parton Model, Vol. 1, Elsevier Science Ltd., 1984

1984
[8]

V.N.Gribov, Space-time description of the hadron interaction at high energies, Lecture given at the VIII Winter Petersburg NPI School, 1973, arXiv:hep- ph/0006158

arXiv 1973

Show all 17 references

[9]

Fadin, E.A

V.S. Fadin, E.A. Kuraev, L.N. Lipatov, On the Pomeranchuk Singularity in Asymptotically Free Theories, Phys. Lett. B. 60 (1975) 50

1975
[10]

Kuraev, L.N

E.A. Kuraev, L.N. Lipatov, V.S. Fadin, The Pomeranchuk singularity in non- abelian gauge theories, Sov. Phys. JETP 45 (1977) 199

1977
[11]

Balitsky, L.N

I.I. Balitsky, L.N. Lipatov, The Pomeranchuk Singularity in Quantum Chromo- dynamics, Sov. J. Nucl. Phys. 28 (1978) 822

1978
[12]

Balitsky, Operator expansion for high-energy scattering, Nucl

I. Balitsky, Operator expansion for high-energy scattering, Nucl. Phys. B 463, 99 (1996)

1996
[13]

Kovchegov, Small-x F(2) structure function of a nucleus including multiple pomeron exchanges, Phys

Y.V. Kovchegov, Small-x F(2) structure function of a nucleus including multiple pomeron exchanges, Phys. Rev. D 60, 034008 (1999)

1999
[14]

Polyakov, A Similarity Hypothesis in the Strong Interactions

A.M. Polyakov, A Similarity Hypothesis in the Strong Interactions. II. Cascade Production of Hadrons and Their Energy Distribution Associated withe +e− Annihilation, Sov. Phys. JETP 33 (1971) 850

1971
[15]

Nekrasov, High-energy Coulomb scattering of spatially extended particles, Nucl

M.L. Nekrasov, High-energy Coulomb scattering of spatially extended particles, Nucl. Phys. A 1058 (2024) 123068

2024
[16]

Iancu, L

E. Iancu, L. McLerran, Saturation and universality in QCD at small x, Phys. Lett. B 510 (2001) 145

2001
[17]

Iancu, R

E. Iancu, R. Venugopalan, The Color glass condensate and high-energy scattering in QCD, in R.C.Hwa and X.N.Wang (Eds.), Quark-Gluon Plasma 3, Singapore, World Scientific, 2004, pp. 249-363, hep-ph/0303204 12

2004 arXiv

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.

[1] [1]

Feynman, Photon-hadron Interactions, W.A

R.P. Feynman, Photon-hadron Interactions, W.A. Benjamin 1972

1972

[2] [2]

Gribov, L.N

V.N. Gribov, L.N. Lipatov, Deep inelastice −pscattering in perturbation theory, Sov. J. Nucl. Phys. 15 (1972) 438

1972

[3] [3]

Dokshitzer, Calculation of the Structure Functions for Deep Inelastic Scat- tering ande +e− Annihilation by Perturbation Theory in Quantum Chromody- namics, Sov

Yu.L. Dokshitzer, Calculation of the Structure Functions for Deep Inelastic Scat- tering ande +e− Annihilation by Perturbation Theory in Quantum Chromody- namics, Sov. Phys. JETP 46 (1977) 641

1977

[4] [4]

Altarelli, G

G. Altarelli, G. Parisi, Asymptotic Freedom in Parton Language, Nucl. Phys. B 126 (1977) 298

1977

[5] [5]

L. V. Gribov, E. M. Levin, M. G. Ryskin, Semihard Processes in QCD, Phys. Rept. 100 (1983) 1

1983

[6] [6]

Mueller, J.Qiu, Gluon Recombination and Shadowing at Small Values ofx, Nucl

A.H. Mueller, J.Qiu, Gluon Recombination and Shadowing at Small Values ofx, Nucl. Phys. B268 (1986) 427. 11

1986

[7] [7]

B. L. Ioffe, V. A. Khoze and L. N. Lipatov, Hard Processes: Phenomenology, Quark-Parton Model, Vol. 1, Elsevier Science Ltd., 1984

1984

[8] [8]

V.N.Gribov, Space-time description of the hadron interaction at high energies, Lecture given at the VIII Winter Petersburg NPI School, 1973, arXiv:hep- ph/0006158

arXiv 1973