---
title: 'Sak''s Scenario: LR vs SR Criticality'
url: https://www.emergentmind.com/topics/sak-s-scenario
type: topic
---

# Sak's Scenario: LR vs SR Criticality

Searching arXiv for recent and foundational papers on Sak’s scenario in long-range critical phenomena.
Sak’s scenario is a renormalization-group criterion for the crossover between long-range (LR) and short-range (SR) critical behavior in statistical models with pair interactions decaying as \(J(r)\sim r^{-(d+\sigma)}\). In its standard form, the crossover threshold is predicted to occur at \(\sigma_* = 2-\eta_{\mathrm{SR}}\), so that the LR anomalous-dimension relation \(\eta=2-\sigma\) matches continuously onto the SR value \(\eta_{\mathrm{SR}}\) at the boundary. In the contemporary literature, the phrase denotes both this continuity-based criterion itself and the broader claim that the LR–SR transition is governed by a shifted boundary rather than by the older Fisher–Ma–Nickel picture with \(\sigma_*=2\). Recent arXiv work treats Sak’s scenario as an active point of controversy rather than a settled theorem across all dimensions and models: large-scale simulations in one-dimensional self-avoiding Lévy flights support it [2507.08092], whereas recent two-dimensional studies of LR Ising, XY, Heisenberg, and percolation models argue against it and advocate \(\sigma_*=2\) instead [2512.04805].

## 1. Definition and theoretical content

Sak’s scenario concerns models with power-law interactions of the form
\[
J(r)\sim \frac{1}{r^{d+\sigma}},
\]
where \(d\) is spatial dimension and \(\sigma>0\) controls interaction range. Small \(\sigma\) corresponds to slower decay and stronger LR effects; larger \(\sigma\) corresponds to faster decay and a tendency toward SR behavior [2507.08092]. In this setting one expects three regimes: a mean-field regime for sufficiently small \(\sigma\), an intermediate genuinely LR critical regime, and an SR regime above a crossover value \(\sigma_*\) [2507.08092].

The historical competing picture reviewed in the recent literature is the Fisher–Ma–Nickel scenario, in which the LR anomalous dimension remains
\[
\eta=2-\sigma
\]
throughout the LR region up to \(\sigma_*=2\), producing a discontinuous jump in \(\eta\) when the system crosses to the SR universality class [2512.04805]. Sak’s refinement shifts the boundary to
\[
\sigma_* = 2-\eta_{\mathrm{SR}},
\]
and thereby enforces continuity of \(\eta\) at the crossover [2507.08092]. In this view, the LR formula \(\eta=2-\sigma\) applies only up to \(\sigma_*\), and at \(\sigma=\sigma_*\) it matches the SR value \(\eta_{\mathrm{SR}}\) [2507.08092].

For the one-dimensional self-avoiding-walk problem studied through the \(n\to 0\) limit of the \(O(n)\) model, the SR exponents are exactly
\[
\nu_{\mathrm{SR}}=1,\qquad \gamma_{\mathrm{SR}}=1.
\]
Assuming hyperscaling,
\[
\gamma=(2-\eta)\nu,
\]
this implies
\[
\eta_{\mathrm{SR}}=1,
\]
and hence Sak’s criterion predicts
\[
\sigma_*=2-\eta_{\mathrm{SR}}=1
\]
in that case [2507.08092].

A central reason the criterion matters is that the location of \(\sigma_*\) determines which universality class controls critical behavior for a given interaction range. This is especially consequential in low-dimensional systems, where finite-size effects and strong corrections to scaling can make the asymptotic boundary difficult to identify numerically [2507.08092].

## 2. Historical dispute and competing interpretations

Recent work frames Sak’s scenario as one branch of a longer dispute about how LR fixed points give way to SR ones. The continuity-based argument is regarded as elegant because it avoids a jump in \(\eta\), but it is also criticized for extrapolating perturbative information beyond its nominal range of control [2512.04805]. In the two-dimensional review by Wang and collaborators, the “bare” continuum description is written as
\[
\beta H= \int\frac{\mathrm d^dq}{(2\pi)^d}\left(\frac t2+\frac{K_2}{2}q^2+K_\sigma q^\sigma\right) \boldsymbol{\Psi}(\boldsymbol q)\cdot \boldsymbol{\Psi}(-\boldsymbol q) +\int \mathrm d^dr\, |\boldsymbol{\Psi}(\boldsymbol r)|^4 ,
\]
and Sak’s argument is characterized as the claim that renormalization can render the nonanalytic \(q^\sigma\) term irrelevant before \(\sigma\) reaches \(2\), so that the analytic \(q^2\) term takes over [2512.04805].

The same paper emphasizes that the perturbative expansion
\[
\eta = 2-\sigma + O(\epsilon^3),\qquad \nu = \frac{1}{\sigma}+\frac{4}{\sigma d}\frac{n+2}{n+8}\epsilon + O(\epsilon^2),
\]
with \(\epsilon=2\sigma-d\ge 0\), is controlled near \(\sigma=d/2\), not near \(\sigma=2\) [2512.04805]. This suggests that using the continuity of \(\eta\) alone to infer the full crossover structure may be too strong. A plausible implication is that Sak’s scenario is best viewed as a specific RG hypothesis about the fate of the LR kinetic term, rather than as a purely kinematic matching condition.

The contemporary literature also distinguishes Sak’s scenario from a different continuity picture associated with Picco. In the four-scenario taxonomy articulated in the 2D critique, one possibility is \(\sigma_*=2\) with a smooth crossover of exponents near \(2\); another is \(\sigma_*=2\) with a weak but genuine discontinuity between \(\sigma=2\) and any \(\sigma>2\) [2512.04805]. That latter position is the one advocated in the recent 2D Monte Carlo study, and it is presented explicitly as an alternative to Sak’s criterion.

## 3. One-dimensional evidence in favor

The strongest recent numerical case for Sak’s scenario comes from the study of one-dimensional self-avoiding Lévy flights by Angelini, Parisi, Picco, and Ricci-Tersenghi [2507.08092]. The model is obtained from the \(n\to 0\) limit of the long-range \(O(n)\) model, using the de Gennes mapping from high-temperature graphs to a single open self-avoiding path [2507.08092]. With long-range couplings, that path becomes a self-avoiding Lévy flight whose jump tail is
\[
P(r)\sim r^{-(1+\sigma)}.
\]

The authors simulate on an infinite one-dimensional lattice and define the SAW exponents through
\[
\langle \log R_N \rangle \sim \nu_{\rm{LR}} \log N, \qquad c_N \sim \mu^N\, N^{\gamma-1},
\]
with generating function
\[
G (R, z) = \sum_{N=0}^{\infty} c_N (R) z^N,
\]
and susceptibility
\[
\chi(z)  =\sum_{R} G (R, z) = \sum_{N=0}^{\infty} c_{N} z^N.
\]
At critical fugacity \(z_c=\mu^{-1}\),
\[
\chi(z) \sim (z_c -z)^{-\gamma} \quad \text{for} \quad z \nearrow z_c
\]
[2507.08092].

The upper critical dimension of Lévy-SAW is stated as
\[
d_\text{crit} = 2\sigma,
\]
so in one dimension the mean-field–LR boundary is at \(\sigma=d/2=1/2\), while the disputed LR–SR boundary is the one addressed by Sak’s criterion [2507.08092]. Their anomalous dimension is extracted from
\[
G(R)\sim R^{1-\eta},
\]
and at \(\sigma=1\) they report logarithmic corrections of the form
\[
G(R) = R^{1-\eta} \cdot \frac{a}{b + \log(R)}.
\]
This is presented as a hallmark of marginal crossover behavior precisely at the Sak boundary [2507.08092].

Numerically, the evidence is direct for \(\eta\). For \(\sigma<1\), the data agree very well with
\[
\eta = 2-\sigma.
\]
At \(\sigma=1\), the estimate is
\[
\eta = 1.011_{-0.062}^{+0.042},
\]
consistent with \(\eta_{\mathrm{SR}}=1\), and for \(\sigma>1\), \(\eta\) remains close to \(1\) [2507.08092]. The transition is therefore continuous in \(\eta\), as Sak’s scenario requires.

The paper also argues that \(\gamma\) and \(\nu\) approach the SR values \(1\) for \(\sigma>1\), though with strong finite-size drift. Their correction-to-scaling analysis is central: in the SR regime, starting from an effective action containing both LR and SR kinetic terms,
\[
\Gamma_k[\phi] = \int d^{d}x \left\{ Z_{\sigma} \partial_{\mu}^{\frac{\sigma}{2}} \phi_{i} \partial_{\mu}^{\frac{\sigma}{2}} \phi_{i} + Z_2 \partial_{\mu} \phi_{i} \partial_{\mu} \phi_{i} + U_k(\rho) \right\},
\]
the flow of the renormalized LR coupling is
\[
\partial_t \bar{J}_\sigma = (\sigma - 2) \bar{J}_\sigma + \eta_2 \bar{J}_\sigma,
\]
yielding
\[
\sigma^* = 2-\eta_2.
\]
In the SR regime,
\[
\partial_t \bar{J}_\sigma = (\sigma - \sigma^*) \bar{J}_\sigma,
\qquad
\bar{J}_\sigma (N) = \bar{J}_\sigma (0)N^{(\sigma - \sigma^*)},
\]
so the correction exponent is predicted to be
\[
\Delta=\sigma-\sigma^*.
\]
At \(\sigma=1.5\), they find \(\omega=0.533(1)\) from \(\gamma\) and \(\omega=0.417(1)\) from \(\nu\), both close to \(\sigma-\sigma^*=0.5\) [2507.08092]. This is presented as strong support for Sak’s scenario and as an explanation of why earlier simulations could misidentify a crossover nearer \(\sigma=2\).

## 4. Quantitative relations and scaling structure

In the one-dimensional Lévy-SAW setting, the paper articulates Sak’s scenario through a set of linked scaling relations [2507.08092]. Assuming hyperscaling, one has
\[
\gamma=(2-\eta)\nu.
\]
Since \(\eta=2-\sigma\) on the LR side and \(\eta_{\mathrm{SR}}=1\) on the SR side, the crossover relation becomes
\[
\gamma =
\begin{cases}
\sigma \,\nu_{\rm{LR}}, & \sigma < \sigma^{*}, \\
\nu_{\rm{LR}}, & \sigma > \sigma^{*}.
\end{cases}
\]
The paper reports that direct measurements of \(\eta\) are consistent with \(2-\gamma/\nu_{\rm LR}\), supporting the hyperscaling interpretation [2507.08092].

The same study also situates Sak’s scenario relative to Flory theory. In 1D Lévy-SAW, Flory predicts
\[
\nu_\text{LR}= \frac{3}{1+\sigma},
\]
and in the SR case gives \(\nu_{SR}=1\), which suggests a smooth approach to SR only at \(\sigma=2\) [2507.08092]. The Monte Carlo results instead indicate that Flory theory is numerically close but not exact. At \(\sigma=2/3\), for example, the asymptotic estimates are
\[
\gamma = 1.183^{+0.005}_{-0.009},\qquad
\nu_{\rm{LR}} = 1.778^{+0.004}_{-0.009},
\]
slightly below the Flory value \(\nu_{\rm Flory}=1.8\) [2507.08092]. This matters because Sak’s scenario is not merely a claim about \(\sigma_*\); it also implies that the LR branch terminates at \(\sigma^*\) rather than evolving continuously all the way to \(\sigma=2\).

The one-dimensional evidence therefore supports a specific asymptotic synthesis: mean-field for \(\sigma<1/2\), genuinely LR criticality for \(1/2<\sigma<1\), and SR criticality for \(\sigma>1\), with logarithmic corrections and slow crossover at \(\sigma=1\) [2507.08092].

## 5. Critiques and two-dimensional counterclaims

The most explicit recent critique appears in the 2D study “On Sak’s criterion for statistical models with long-range interaction” [2512.04805]. That work examines Ising, XY, Heisenberg, and percolation models with interactions decaying as \(1/r^{2+\sigma}\), and argues for a unified boundary
\[
\sigma_*=2
\]
across all studied systems [2512.04805]. In this account, continuity of \(\eta\) is not taken as compulsory, and the natural threshold is where the nonanalytic \(q^\sigma\) term ceases to dominate over the analytic \(q^2\) term.

For the 2D Ising model, Sak’s criterion predicts
\[
\sigma_*=\frac{7}{4}=1.75
\]
because \(\eta_{\rm SR}=1/4\) [2512.04805]. The critique argues that if Sak were correct, then all \(\sigma>7/4\), including \(\sigma=1.875\) and \(\sigma=2\), should already display SR universal quantities and \(\eta=1/4\). Their Monte Carlo data instead show persistent LR behavior at \(\sigma=2\). The Fortuin–Kasteleyn critical polynomial yields
\[
R_{p,c}=-0.342(5)\ \text{at}\ \sigma=1.75,\quad
-0.207(9)\ \text{at}\ \sigma=1.875,\quad
-0.065(8)\ \text{at}\ \sigma=2,
\]
whereas for \(\sigma=2.2\) and \(2.5\) it is consistent with the SR value \(0\) [2512.04805]. Their anomalous-dimension estimates are
\[
\eta=0.335(4)\ \text{at}\ \sigma=1.75,\quad
0.293(3)\ \text{at}\ \sigma=1.875,\quad
0.273(3)\ \text{at}\ \sigma=2,\quad
0.250(1)\ \text{at}\ \sigma=2.5,
\]
again interpreted as evidence that \(\sigma=2\) still belongs to the LR side [2512.04805].

The critique extends beyond Ising. It argues that Sak’s criterion is conceptually problematic in 2D XY because \(\eta_{\mathrm{SR}}\) varies continuously in the BKT phase, problematic in 2D Heisenberg because no finite-temperature SR critical point exists, and problematic in percolation because \(\eta_{\mathrm{SR}}<0\) in \(2<d<6\), which would imply \(\sigma_*>2\) [2512.04805]. This suggests that the universality of Sak’s formula across model classes is contested.

A plausible interpretation is that the contemporary debate has shifted from whether Sak’s scenario is mathematically elegant to whether it is universally realized. The 1D Lévy-SAW results argue that it is, at least there [2507.08092]. The 2D Monte Carlo program argues that it is not, and that the apparent support in earlier work may reflect finite-size ambiguity near a marginal boundary at \(\sigma=2\) [2512.04805].

## 6. Methodological significance and current status

A notable feature of the present literature is that the dispute is driven less by purely formal RG arguments than by numerical diagnostics sensitive to slow crossover. In the one-dimensional supporting study, the key signatures are continuity of \(\eta\), logarithmic corrections at \(\sigma=1\), and correction exponents compatible with \(\Delta=\sigma-\sigma^*\) in the SR regime [2507.08092]. In the two-dimensional critique, the decisive diagnostics are the FK critical polynomial \(R_p\), the anomalous dimension \(\eta\), and large-\(L\) finite-size scaling up to \(L=8192\) [2512.04805]. Both sides emphasize that near the crossover, finite-size effects are unusually strong.

The present status is therefore mixed. In one dimension, the recent Monte Carlo evidence is presented as a numerical resolution of the LR–SR crossover controversy in favor of Sak’s scenario, with
\[
\sigma^*=1
\]
and continuous \(\eta\) [2507.08092]. In two dimensions, recent simulations argue that the crossover is instead at
\[
\sigma_*=2
\]
for all studied models, with the boundary point itself on the LR side and a weak but genuine discontinuity between \(\sigma=2\) and any \(\sigma>2\) [2512.04805].

It would therefore be misleading to treat “Sak’s scenario” as either an obsolete idea or an established universal law. The more accurate characterization is that it remains a central organizing hypothesis in LR critical phenomena: supported in some settings, disputed in others, and still structurally important because it ties the LR–SR boundary to the SR anomalous dimension rather than to the naive decay exponent alone. This suggests that its lasting significance lies not only in the formula \(\sigma_*=2-\eta_{\mathrm{SR}}\), but also in the methodological insistence that crossover boundaries may be determined by renormalized rather than bare scaling data.

Source: https://www.emergentmind.com/topics/sak-s-scenario