{"id":"f6b4ceb7-e3d0-4368-ad55-f948f25cb111","arxiv_id":"2505.10396","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Lattice simulations show that on T2 x R2 with twisted boundary conditions, the SU(2) vacuum is a Poissonian 2D gas of fractional instantons, with string tension proportional to the gas density and approaching the infinite-volume value.","lead":"This paper tests whether SU(2) gluon fields on a small twisted two-torus behave as a dilute gas of half-instanton vortices whose density and string tension grow together. The Monte Carlo data match the semiclassical picture up to torus sizes around 0.6 fermi, suggesting a controlled route from weak coupling to confinement.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Density entering the sigma–rho relation is measured after Wilson flow at tau=0.65 fm; Fig. 19 shows strong flow-time dependence, so the claimed proportionality and semiclassical scaling may depend on the chosen smearing radius.","rationale":"The reader's weakest assumption identified exactly the flow-time and threshold sensitivity of the density extraction. My pass agrees and sharpens it: the paper's own Figs. 19 and 20 show that the measured density and the sigma/rho ratio depend on the flow time, and the Poissonian tests cannot exclude a systematic thinning of the gas. This justifies the CONDITIONAL verdict already given; no adjustment of the reader's verdict is needed, but the concrete test above would resolve the residual uncertainty. I did not find an independent internal inconsistency in the Poisson-gas formulas, and I credit the paper for its explicit discussion of the flow-time limitation and for the smooth-configuration checks of the center-vortex behavior. The remaining risk is quantitative calibration of the identification pipeline, not the existence of vortex-like fractional instantons on the twisted torus.","tokens_in":35889,"tokens_out":9337,"duration_ms":98061,"concrete_test":"Generate synthetic ensembles on the same Ls = 6, 8, 10 lattices by placing a known 2D Poisson gas of the numerically computed FI/AFI solutions (Sec. 2.2) at a chosen input density, apply exactly the Wilson flow at tgf values 4, 8, 15, and run the identification algorithm of Sec. 3.5 to compare reconstructed vs input density as a function of tgf and pair distance. If the fractional loss and pairing-distance bias are large and density-dependent, use the measured response function to correct the Monte Carlo densities; then re-fit Eq. (4.2) and Fig. 13 and the Fig. 16 correlation. The concern is settled if the corrected exponent remains consistent with 11/3 and the sigma/rho slope moves to the TAVA expectation of 2 plus the finite-size corrections documented in Sec. 2.4.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing quantitative input is the 2D density rho_2D of FI/AFI extracted from flowed configurations. Sec. 3.5 counts peaks that survive a fixed 20%-qfrac threshold after Wilson flow at a fixed physical smearing radius tau ≈ 0.65 fm (Tab. 1). The paper itself states in Sec. 4.4 that rho_2D cannot be measured at zero flow time and that the density decreases monotonically with flow time; Fig. 19 shows a large decrease between tgf ≈ 2 and tgf ≈ 20, and the text of Sec. 4.4 says the sigma/rho ratio approaches the TAVA value slightly above 2 at smaller flow times, whereas the value used in Fig. 16 is the larger value at tgf = 15. The Poissonian number and nearest-neighbour tests in Sec. 4.1 do not protect against this, because distance- or sign-dependent thinning of a Poisson process is again a Poisson process, and the opposite-sign deficit in Fig. 11 already indicates such thinning. Consequently, the agreement of the surviving density with the semiclassical formula in Sec. 4.2 and the linear relation sigma ≈ 2.89 rho_2D could be an artifact of the selected flow time rather than evidence that the physical FI gas generates the string tension. The central claim would be secure only if the same scaling and proportionality survive a systematic zero-flow extrapolation or a calibrated correction for flow-induced annihilation and fusion.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies SU(2) Yang-Mills theory on a T2 x R2 geometry with one unit of 't Hooft twist in the small torus, monitoring the system as the torus size ls is varied. For small ls, the authors argue that the long-distance physics is described by a dilute, Poissonian two-dimensional gas of vortex-like fractional instantons (FI) and anti-instantons (AFI). They present Monte Carlo data at several lattice couplings and torus sizes, an identification algorithm based on Wilson-flowed topological charge density peaks, and tests of the semiclassical predictions for the density's beta-dependence and for the relation between the string tension and the 2D density. The main quantitative claims are that the diluteness follows the semiclassical form A(Ls) beta^2 exp(-beta S_L/4) with A(Ls) ~ Ls^{3.45}, and that the effective string tension is proportional to the FI density with slope 2.89, approaching the infinite-volume value as ls grows. The paper also includes preliminary results for SU(3) and SU(4) and a discussion of the transition toward larger torus sizes.","tokens_in":1838,"tokens_out":1962,"duration_ms":74870,"significance":"If the central claims hold, the paper provides one of the most quantitative semiclassical descriptions of confinement on a T2 x R2 geometry, connecting the dilute FI gas to the string tension and to the fractional-instanton liquid picture. The work is strengthened by an extensive ensemble scan, a detailed and transparent identification methodology, an explicit study of flow-time dependence in Sec. 4.4, and a candid statement in Sec. 6 that the results do not by themselves prove that vortices cause confinement in the infinite-volume theory. The analysis also demonstrates the stability of individual FI solutions under gradient flow and validates the Poissonian nature of the gas at low density using number counts, nearest-neighbor distributions, and topological-charge histograms. The main limitations are that the absolute density is not predicted (the normalization and exponent are fitted) and that the sigma-rho proportionality is established at a single fixed physical smearing radius, with a slope that moves toward the TAVA value 2 at smaller flow times.","major_comments":[{"comment":"The claim in Sec. 6 that \"The density of objects matches closely the predictions of the theory\" is stronger than what the analysis establishes. The beta-dependence of the diluteness is tested by fitting the prefactor A(Ls) in Eq. (4.2), and the continuum scaling is then checked by fitting the exponent alpha in Eq. (4.3) to the same fitted A(Ls) values. Thus the absolute density is not a parameter-free prediction, and the agreement with the exponent 11/3 is a consistency check of the functional form rather than a derivation of the density. Please revise the wording to state explicitly that only the shape of the beta-dependence and the scaling exponent are tested, and that the normalization is a fitted quantity.","section":"Sec. 4.2, Eqs. (4.2)-(4.3)"},{"comment":"The proportionality sigma = c rho_2D with c = 2.89 is obtained from data at a fixed physical smearing radius tau = 0.65 fm (Tab. 1), but Sec. 4.4 shows that the density decreases strongly with flow time while the ratio sigma/rho approaches the TAVA value slightly above 2 at smaller tgf. This means the slope c in Fig. 16 is not a property of the physical FI gas alone; it depends on the chosen flow time. The authors should either provide a systematic zero-flow extrapolation of rho (or a quantitative correction for flow-induced annihilation) and demonstrate that c is stable, or explicitly present the sigma-rho relation as a flow-time-dependent effective quantity. Without this, the statement that \"the very same effective string tension is proportional to the density of fractional instantons\" is not supported at the quoted precision.","section":"Secs. 4.3-4.4, Figs. 16 and 20"},{"comment":"The error bars on the density rho_2D in Fig. 14 and in the sigma-rho plot include only the systematic variation of the identification thresholds, not the much larger flow-time dependence documented in Fig. 19. Since the density is a central input to the scaling plot and to the sigma-rho relation, the quoted uncertainties underestimate the true systematic error. Please add a flow-time systematic to the quoted errors, or justify that the chosen tau lies in a plateau region across the full beta and Ls range used in the analysis.","section":"Sec. 4.4 and Fig. 14"}],"minor_comments":[{"comment":"The lattice spacing for beta = 2.45 is listed as 0.985 fm, which is an order of magnitude larger than the neighboring values (0.1163 and 0.0819 fm); this appears to be a typo for 0.0985 fm. Please correct.","section":"Tab. 1"},{"comment":"The fitting form for log(S2(r)/S2(0)) contains the term F·D r^3 in the numerator with a denominator 1.0 + C r + D r^2; the notation is confusing because F and D appear as independent parameters. Please define the intended functional form more clearly.","section":"Sec. 2.2, Eq. (2.13)"},{"comment":"The x-axis of Fig. 19 is labelled tgf but the text refers to flow times in different units; please specify in the caption whether this is the lattice gradient-flow time or a physical quantity, and what range in physical units (tau) corresponds to the displayed values.","section":"Fig. 19"},{"comment":"The word \"excentricity\" is used in the text and in Eq. (3.4); the standard spelling is \"eccentricity\". Please correct throughout.","section":"Sec. 3.4"},{"comment":"The value of the fitted coefficient A is written as \"3.91(10) 106\" in the text; please format it as 3.91(10) x 10^6 so that the exponent is unambiguous.","section":"Sec. 4.2, Fig. 13"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis is a substantial lattice study of SU(2) Yang-Mills on T2 x R2 with one unit of 't Hooft twist. The paper carefully verifies that in the small-torus regime the vacuum looks like a dilute 2D gas of vortex-like fractional instantons: number counts are Poissonian, nearest-neighbour distributions match the gas prediction, the beta-dependence of the density follows the semiclassical exponential with a fitted prefactor, and the prefactor exponent, 3.45(2), is close to the 11/3 RG prediction. The authors also document the transition to non-dilute behaviour around ls ~ 0.6 fm and give some SU(3)/SU(4) outlook. That is a real piece of work with a lot of Monte Carlo data behind it.\n\nThe soft spots are where the reader and the stress-test point. The density is defined after Wilson flow at a fixed smearing radius tau ~ 0.65 fm. The paper itself shows that rho_2D falls strongly with flow time and that sigma/rho_2D decreases from about 2.9 at tgf = 15 to slightly above 2 at smaller times. So the measured proportionality constant is not universal; it is flow-time dependent. The claim that sigma ~ 2.89 rho_2D is true only for this flow prescription, and the suggestion that the TAVA value 2 is recovered at zero flow is an extrapolation that is not actually carried out. To make the roadmap claim solid, the authors need a systematic zero-flow extrapolation of the density, a calibrated annihilation correction, or at least a demonstration that the scaling and proportionality are stable over a range of flow times.\n\nAlso, the semiclassical agreements are consistency checks rather than parameter-free predictions: A(Ls) is free, the exponent is fitted, and the Poisson and distance tests use the measured mean density as input. That is fine, but it should be described as confirmation of the form, not as derivation of the density. The fit excludes beta <= 2.5, and the string tension is extracted from Creutz ratios in a rather narrow window, adding systematic uncertainty. There is a typo in Eq. (2.30): the exponent should be -3 pi^2 beta / 11 rather than -3 beta / (11 pi^2); I assume that is a typesetting slip.\n\nThe paper is honest about limits, explicitly stating that the results do not prove vortices underpin confinement in the infinite-volume theory. That is the right attitude. The work deserves a serious referee and, I think, publication after revision. The main ask is to address the flow-time dependence of the density and to reframe the sigma-rho relation accordingly. I would bring it to a reading group.","headline":"A serious lattice study that verifies the semiclassical 2D gas picture on T2 x R2, but the density–string tension link is flow-time dependent and the headline claim goes a bit beyond the evidence.","tokens_in":36798,"tokens_out":3786,"would_cite":true,"duration_ms":37025,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper claims that a dilute gas of vortex-like fractional instantons generates the confining string tension in SU(2) Yang-Mills on a twisted two-torus, with the measured string tension proportional to the gas density and approaching…","keywords":["confinement","fractional instantons","center vortices","twisted boundary conditions","lattice Yang-Mills","string tension","semiclassical approximation","gradient flow"],"falsifier":"Measure the object density with a flow-independent method (for example adjoint zero-mode filtering or overimproved cooling) and compare it with the Wilson-flow identification; if the zero-flow-time density is not proportional to the string tension with a slope near 2, or if the topological charge distribution deviates from the Poisson prediction $P(Q)=I_{2Q}(2p)/\\cosh(2p)$, the central claim fails.","tokens_in":35632,"feed_emoji":"🌀","tokens_out":8366,"duration_ms":70502,"temperature":0.7,"pith_summary":"This paper tries to show that confinement in SU(2) Yang-Mills can be traced to a dilute two-dimensional gas of vortex-like fractional instantons when two space-time directions are compactified on a small twisted torus. At torus sizes below about 0.6 fm, the Monte Carlo ensembles match the semiclassical prediction of a Poissonian gas whose density is set by the one-instanton weight and grows with the torus size. The measured string tension rises linearly with that density, with a slope close to the value 2 that a 2D center-vortex gas predicts, and approaches the infinite-volume string tension as the torus grows. The authors read this as evidence that fractional instantons, not ordinary $Q=1$ instantons, generate the confining force.","feed_headline":"Fractional instantons can carry the SU(2) string tension","feed_subtitle":"On a twisted two-torus, the measured string tension tracks the gas density and approaches the infinite-volume value.","key_machinery":"The central object is the vortex-like fractional instanton of topological charge $Q=1/2$ on a twisted $T^2\\times \\mathbb{R}^2$: a self-dual classical solution localized in the large plane, behaving as a $\\mathbb{Z}_2$ center vortex, with action half that of an instanton. The macroscopic description is a two-dimensional Poisson gas of these objects and their anti-instantons. The load-bearing identity is the thin-abelian-vortex approximation (TAVA), $\\sigma = 2\\rho_{2D}$, which turns the gas density into a string tension; the paper tests it through the diluteness $D = L_s^2\\rho_{2D}$ and its predicted scaling $D = A(L_s)\\beta^2 e^{-\\pi^2\\beta}$ with $A(L_s)\\sim L_s^{11/3}$.","core_discovery":"The paper claims that on $T^2\\times \\mathbb{R}^2$ with one unit of 't Hooft twist, the long-distance vacuum of SU(2) Yang-Mills is a Poissonian 2D gas of fractional instantons ($Q=1/2$) and anti-instantons, and that the string tension in the large plane is proportional to the gas density, with the data giving $\\sigma \\approx 2.89\\, \\rho_{2D}$, close to the thin-abelian-vortex prediction $\\sigma = 2\\rho_{2D}$. The density follows the semiclassical diluteness formula $D = L_s^2\\rho_{2D} = A(L_s)\\, \\beta^2 e^{-\\pi^2 \\beta}$, with $A(L_s)\\sim L_s^{3.45(2)}$, near the renormalization-group exponent $11/3$, and the physical density scales as $(l_s\\Lambda)^{5/3}$. Combined with the fact that the string tension at $l_s\\sim 0.8$ fm is already close to the infinite-volume value, the paper concludes that the confinement property of the infinite-volume theory has its origin in fractional instantons.","pith_inferences":["If the fractional-instanton gas is the true confining mechanism, the ratio $\\sigma/\\rho_{2D}$ should tend to 2 for Wilson loops much larger than the instanton size; the measured 2.89 is likely a finite-loop and flow-time artifact that can be checked with larger smeared loops.","The flow-time dependence of the density implies that the zero-flow-time density is higher than the quoted value; an extrapolation of $\\rho_{2D}(t_{gf})$ to $t_{gf}\\to 0$ at fixed physical smearing radius should preserve the proportionality $\\sigma\\propto \\rho_{2D}$, providing a sharper test.","A testable prediction of the Poissonian hypothesis is that the topological charge distribution should follow $P(Q)=I_{2Q}(2p)/\\cosh(2p)$ with $p=\\rho_{2D}A/2$, so high-precision histograms of $Q$ can falsify the interpretation without relying on peak identification.","Because the construction relies on 't Hooft twist, the same gas should reappear in other compactifications (for instance as calorons or monopoles on $S^1\\times\\mathbb{R}^3$); if fractional instantons are universal, the different semiclassical pictures are related by geometry, which could be tested by interpolating between tori."],"forward_implications":["At $l_s$ below about 0.6 fm, the fractional-instanton gas is dilute and Poissonian; number counts, neighbour-distance distributions, and the topological charge distribution match a free 2D gas with only the mean density as input.","The string tension in the $T^2\\times\\mathbb{R}^2$ geometry is set by the gas density; $Q=1$ instantons, though present, do not contribute to the string tension.","The density and diluteness scale with $l_s$ and $\\beta$ according to the semiclassical weight and the beta function, so the picture survives the continuum limit.","As $l_s$ grows beyond roughly 0.6 fm the gas becomes non-dilute, the size of the fractional instantons decouples from $l_s$ and is set by the mean inter-object distance, and the string tension approaches its infinite-volume value.","Preliminary SU(3) and SU(4) results show the expected hierarchy of fractional charges, with $Q=1/N$ objects dominating the density."],"supporting_citations":[{"why":"Defines the $Q=1/2$ vortex-like fractional instanton on $T^2\\times\\mathbb{R}^2$, the building block of the gas.","marker":"[48]"},{"why":"Introduces twisted boundary conditions and 't Hooft flux, the setting that makes fractional instantons the natural semiclassical degrees of freedom.","marker":"[1]"},{"why":"Provides the previous semiclassical gas analysis on the twisted torus whose Poissonian density formulas and string-tension connection are adapted here.","marker":"[22]"},{"why":"Proposes the relation between fractional-instanton gas density and the string tension and the fractional instanton liquid picture that motivates the large-$l_s$ interpretation.","marker":"[24]"},{"why":"Supplies the one-loop quantum weight for (fractional) instantons, fixing the $\\beta$-dependence of the gas density.","marker":"[44]"},{"why":"Introduces the gradient flow used to filter ultraviolet noise before identifying topological objects.","marker":"[58]"},{"why":"Provides the modified instanton profile used to fit the peaks in the topological charge density.","marker":"[66]"},{"why":"Gives the lattice spacing scale $a(\\beta)$ used to convert lattice sizes and densities to physical units.","marker":"[53]"}],"fun_headline_variants":["Fractional instanton gas drives SU(2) confinement","Twisted torus shows instantons govern string tension","String tension from density of fractional instantons","Yang-Mills confinement traced to vortex gas on torus","Semiclassical gas sets confinement scale in SU(2)"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The density of the gas is read off from peak heights in the topological charge density after gradient flow at a fixed smearing radius; if the flow systematically removes or merges the very objects that set the string tension, or if the 20% $q_{frac}$ threshold misclassifies noise, both the measured density and the $\\sigma/\\rho$ relation shift.","fun_headline_variants_meta":{"raw":{"variants":["Fractional instanton gas drives SU(2) confinement","Twisted torus shows instantons govern string tension","String tension from density of fractional instantons","Yang-Mills confinement traced to vortex gas on torus","Semiclassical gas sets confinement scale in SU(2)"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00054,"raw_usage":{"total_tokens":2610,"prompt_tokens":988,"completion_tokens":1622,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":604,"completion_tokens_details":{"reasoning_tokens":1544}},"tokens_in":604,"tokens_out":1622,"duration_ms":12359,"temperature":1.0,"reasoning_tokens":1544,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:10:36.681011+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the object density with a flow-independent method (for example adjoint zero-mode filtering or overimproved cooling) and compare it with the Wilson-flow identification; if the zero-flow-time density is not proportional to the string tension with a slope near 2, or if the topological charge distribution deviates from the Poisson prediction $P(Q)=I_{2Q}(2p)/\\cosh(2p)$, the central claim fails.","supporting_citations":[{"cited_title":"’t Hooft,A Property of Electric and Magnetic Flux in Nonabelian Gauge Theories, Nucl","cited_arxiv_id":null,"evidence_quote":"Introduces twisted boundary conditions and 't Hooft flux, the setting that makes fractional instantons the natural semiclassical degrees of freedom."}],"review_version":1}