{"id":"0b1b93cb-0988-40bf-811f-47e18e5a5271","arxiv_id":"2507.18472","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For N=2, the polarized IKKT model is dominated by diverging commuting matrices at small Omega and by the fuzzy sphere at large Omega, with a smooth saddle connection between the two regimes.","lead":"This paper simulates a two-matrix version of the IKKT model to see how a mass-like deformation changes the dominant spacetime configurations. It reproduces the exact partition function and measures new observables that reveal a smooth crossover from the original saddle to a fuzzy sphere, with diverging commuting matrices dominating at weak deformation.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The smooth-connection claim is established only inside a symmetry-restricted ansatz; the full saddle-point equations are not solved, so other saddle branches contributing over the simulated Omega range are not ruled out.","rationale":"The reader's weakest assumption is precisely the load-bearing gap I find: the x1=x2=x3 ansatz is used to derive the asymptotic (13) and the smooth-connection claim, but its completeness is not demonstrated for Omega>0. This matters because the polarized deformation reduces the symmetry, so the Omega=0 reduction to (9) does not automatically justify the same ansatz away from Omega=0. The paper's Monte Carlo results are internally consistent and the partition-function check against localization is strong independent support; the two-peak histograms give evidence for the competing commuting and fuzzy-sphere families, but they do not analytically exclude a third family. I therefore agree with the conditional verdict and do not think the concern escalates it. The proposed numerical saddle scan is a finite, decisive check that would settle whether the ansatz misses any contributing saddle branch.","tokens_in":9527,"tokens_out":18685,"duration_ms":203199,"concrete_test":"Solve the full N=2 saddle-point equations (8) numerically on the 30 real components A_mu = c_{mu,a} sigma_a/2 without imposing x1=x2=x3, for Omega in [0.3,12]. Use continuation from the Omega=0 saddle x_i = 2^{3/4} together with many random Newton starts (e.g., several thousand initial conditions) and count the real solutions with finite action, tracking the branch that reduces to (11). If the branch continuing from Omega=0 is exactly the x1=x2=x3 family, tends to x = 3 Omega/8, and no new real saddles with action comparable to the known two families appear, the concern is resolved. If additional branches appear, or the continuation terminates or bifurcates, the classification in the Discussions and the smooth-connection claim need revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract's central claim is that the Omega=0 saddle is smoothly connected to the large-Omega fuzzy sphere and that the only competing family is the diverging commuting-matrix saddle. For this to be true, the one-parameter family A_a = x sigma_a/2 with x1=x2=x3 must be the relevant fuzzy-sphere branch for all Omega, and no other saddle family can appear in 0.3 <= Omega <= 12. The reduction to (9) is justified at Omega=0 by the full SO(10) x SU(2) symmetry, but the polarized model has only SO(3) x SO(7) x SU(2), so (9) with x1=x2=x3 is an ansatz, not a general parametrization. The paper solves the saddle-point equation only within this ansatz and within the commuting ansatz (14); it does not solve the full 30-dimensional saddle-point problem (8), and the closed-form solution of (12) is deferred to [48]. The exact agreement with the localization partition function checks the partition function, not the geometric classification of saddles. The histograms in Fig. 3 are projections onto log rho3 and log rho7, so an additional saddle with similar extent in these variables would not be visible. Hence 'we were able to identify all the saddle points that contribute' and the smooth-connection statement remain conditional on the unproved ansatz.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports Monte Carlo simulations of the N=2 polarized IKKT matrix model for deformation parameter Omega in the range 0.3 <= Omega <= 12, using a parallel tempering HMC algorithm. The derivative of the log partition function is computed as an expectation value and is found to agree with the SUSY localization result over the whole range, and the small-Omega and large-Omega limits are compared with one-loop and fuzzy-sphere saddle predictions. The authors also measure the extents rho3 and rho7 in the polarized and unpolarized directions, observing a crossover from diverging commuting-matrix configurations at small Omega to a fuzzy-sphere configuration at large Omega. The central claim is that the Omega=0 saddle of the original IKKT model is smoothly connected to the large-Omega fuzzy sphere, while the diverging commuting matrices are responsible for the Omega->0 divergence of the partition function. A toy integral is used to argue that the phenomenon is generic.","tokens_in":9780,"tokens_out":10001,"duration_ms":98028,"significance":"If the central claim holds, the paper clarifies the geometric origin of the recently found Omega->0 singularity and the nature of the intermediate transition in the simplest N=2 case of the polarized IKKT model, and it demonstrates that Monte Carlo simulations can access observables beyond the localization formula. The paper's strengths include the parameter-free validation of the simulation against the localization partition function over a wide Omega range, and the matching of the small-Omega and large-Omega asymptotics to independently derived one-loop and saddle-point expressions. The main caveat is that the smooth-connection statement is derived within a symmetry-restricted ansatz rather than from a complete saddle-point analysis, and no statistical uncertainties are reported for the Monte Carlo data.","major_comments":[{"comment":"The central claim that the Omega=0 saddle is smoothly connected to the large-Omega fuzzy sphere rests on the ansatz x1=x2=x3 in Eq. (9). At Omega != 0 the model has only SO(3) x SO(7) x SU(2) symmetry, so this ansatz is not forced by symmetry, and the full saddle-point equations (8) for the three coefficients xa are not solved. Unless one proves, or verifies numerically, that no other fuzzy-sphere-type branch exists or contributes for 0.3 <= Omega <= 12, the statements that the Omega=0 saddle is 'smoothly connected' to the large-Omega fuzzy sphere and that 'all the saddle points that contribute' were identified are not established. I request a full solution of the three-variable saddle-point equations, or a numerical scan over x1, x2, x3, to rule out additional branches.","section":"Fuzzy sphere saddle, Eqs. (9), (12), (13), and the Discussions paragraph"},{"comment":"No statistical uncertainties are reported for any of the Monte Carlo quantities. The claimed 'complete agreement' with the localization result and the one-loop/saddle predictions cannot be assessed quantitatively without error bars. Please include error bars or a table of statistical errors for all plotted quantities, together with the number of samples and relevant autocorrelation times.","section":"Monte Carlo result, Figs. 1 and 2"},{"comment":"The identification of the left peak in the histograms with the fuzzy-sphere branch that connects to the Omega=0 solution is made by inspection. A quantitative comparison of the peak positions as functions of Omega with the saddle prediction (13) would provide direct evidence for the smooth-connection claim. Without such a comparison, the histograms only show that two metastable families coexist, not that the left peak is on the same analytic branch as the Omega=0 saddle.","section":"Monte Carlo result, Fig. 3"}],"minor_comments":[{"comment":"The notation 'Omega^8/216 E' is ambiguous; the intended expression appears to be Omega^8 / 2^16 times E. Please clarify the typesetting.","section":"Eq. (15)"},{"comment":"The coefficient '9 Omega^2 / 27' in the effective action should be simplified to Omega^2 / 3 for readability, and the same simplification should be applied consistently in Eq. (15).","section":"Eq. (12)"},{"comment":"The text defines the observables as rho3 and rho7 in Eq. (17), but the figure caption refers to R3 and R7. Please unify the notation.","section":"Fig. 2 caption"},{"comment":"The sentence 'The situation simplifies for Omega != 0 since the O(Omega) fermionic mass term in (2) induces the quadratic terms of the fermionic diagonal components' would be clearer if it explained that this renders the fermionic integration Gaussian.","section":"Text below Eq. (4)"}],"recommendation":"major_revision","confidential_remarks":"The paper is well written and the simulations appear carefully executed, but the main geometric claim is conditional on an unproven ansatz and the numerical results lack error estimates. Both issues are fixable within the scope of a letter: the authors can either solve the full three-variable saddle-point equations or provide numerical evidence against other branches, and they can add error bars or a statement about statistical precision. If the authors cannot rule out other saddle branches, they should weaken the 'smoothly connected' and 'all saddle points' claims accordingly. The topic fits the journal's scope, and I see no concerns about novelty or the citation pattern."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"J, here is the quick take on 2507.18472. It is a solid numerical study of the polarized IKKT model at N=2. What is new is the calculation of the observables rho3 and rho7 and the two-branch picture: a fuzzy-sphere branch smoothly connected to the Omega=0 saddle, and a commuting-matrix branch that diverges as 1/Omega and dominates for small Omega. The Monte Carlo matches the localization partition function precisely over a wide range of Omega, which is a strong external check. The histograms show clear double-peak structures and the crossover near Omega~4. The toy-model example (18)-(20) is a nice illustration of how a tiny deformation can create a new dominant saddle at infinity.\n\nThe main soft spot is the geometric classification. The smooth-connection claim is derived from a restricted ansatz, x1=x2=x3=x, and the full 30-dimensional saddle-point equations are not solved. So the statement 'we were able to identify all the saddle points that contribute' is stronger than what is actually proven. The agreement with localization checks the partition function, not the saddle taxonomy. The histograms are projections, so an extra saddle with similar rho3/rho7 would be invisible. That said, the paper is explicit about taking (9) as an ansatz and defers the closed-form solution to a companion paper. The numerical evidence is suggestive and consistent, but the uniqueness claim is conditional. Two more moderate issues: no statistical uncertainties are reported, and no code or data archive is provided, which makes independent verification harder. These are fixable.\n\nWho is this for? People working on matrix-model approaches to spacetime emergence and on deformed IKKT/BMN models. It is a small step from N=2 to large N, but the mechanism it exposes is concrete and the validation is clean. I would send it to a serious referee, not desk reject. The referee should push on the ansatz and ask for error bars and code release. My own verdict: the physics is plausible and the numerics are well executed; I would treat the central claim as likely correct but not fully proven.","headline":"A clean N=2 Monte Carlo study with a strong check against localization; the saddle-classification claim is convincing but rests on an unproven symmetric ansatz.","tokens_in":10298,"tokens_out":2263,"would_cite":true,"duration_ms":22893,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["11.25.-w","02.70.Uu"],"model":"deepseek-v4-flash","headline":"Monte Carlo simulations of the N=2 polarized IKKT matrix model show that the saddle point of the original IKKT model is smoothly connected to the fuzzy-sphere saddle at large deformation parameter Omega, while at small Omega the dominant…","keywords":["IKKT matrix model","polarized IKKT model","fuzzy sphere","commuting matrices","Monte Carlo simulation","parallel tempering","SUSY localization","emergent spacetime"],"falsifier":"Evaluate the full $N=2$ saddle-point equation (8) numerically without imposing $x_1=x_2=x_3$ and without setting $A_I=0$, on a grid of $\\Omega$ in $[0.3,12]$. If any stationary configuration appears whose effective action is comparable to or lower than the two identified families, the claim that all contributing saddle points were captured fails. A direct Monte Carlo check is to start from an asymmetric initial condition near $\\Omega=1$ and see whether a third peak emerges in the $\\log\\rho_3$ or $\\log\\rho_7$ histogram.","tokens_in":9297,"feed_emoji":"🎲","tokens_out":13889,"duration_ms":124647,"temperature":0.7,"pith_summary":"This paper studies the polarized IKKT matrix model, a supersymmetry-preserving mass deformation of the type IIB (IKKT) matrix model, at matrix size $N=2$ over a wide range of the deformation parameter $\\Omega$. Using Monte Carlo simulations with parallel tempering, it reproduces the partition function obtained by supersymmetric localization and then computes observables that localization cannot reach. The central finding is that the saddle point representing the undeformed IKKT model is smoothly connected, as $\\Omega$ grows, to the fuzzy-sphere saddle that dominates at large $\\Omega$; as $\\Omega$ shrinks, the dominant configurations are instead commuting matrices whose entries diverge as $1/\\Omega$. This explains why the partition function diverges as $\\Omega^{-2}$ and why the $\\Omega \\to 0$ limit does not return the finite original IKKT model. The histogram data show a double-peak crossover around $\\Omega \\sim 4$, where the two types of configuration exchange dominance.","feed_headline":"Small deformation of IKKT model is hijacked by commuting matrices","feed_subtitle":"At N=2, the Omega=0 saddle connects to the fuzzy-sphere branch, yet as Omega shrinks, diverging commuting matrices dominate instead.","key_machinery":"The load-bearing object is the effective action $S_{\\mathrm{eff}}(A)=S_b(A)-\\log\\operatorname{Pf}(M(A))$ obtained after integrating out the sixteen fermionic matrices, together with its saddle-point equation $dS_{\\mathrm{eff}}/dA=0$. For $N=2$ the Pfaffian is real and positive semidefinite, so the dominant saddles are real and the sign problem is absent. The argument then rests on two symmetry-reduced ansatze: the fuzzy-sphere ansatz $A_a=x\\,\\sigma_a/2$ with $x_1=x_2=x_3=x$ and $\\sigma_a$ the fundamental $2\\times2$ su(2) generators, which carries the smooth connection from the $\\Omega=0$ saddle to the large-$\\Omega$ fuzzy sphere, the matrix analogue of an ordinary two-sphere; and the commuting ansatz $A_3=x\\,\\sigma_3/2$, $A_{10}=y\\,\\sigma_3/2$, which captures the $O(1/\\Omega)$ diverging configurations responsible for the singularity. Parallel tempering across the two widely separated saddle families is what lets the Monte Carlo simulation sample both peaks with the correct weights.","core_discovery":"After integrating out the fermions, the model is controlled by the effective action $S_{\\mathrm{eff}}(A)=S_b(A)-\\log\\operatorname{Pf}(M(A))$, whose saddle-point equation $dS_{\\mathrm{eff}}/dA=0$ has real solutions for $N=2$ because the Pfaffian is real and positive semidefinite. For the fuzzy-sphere family the paper uses the ansatz $A_a=x\\,\\sigma_a/2$ with $x_1=x_2=x_3=x$; the saddle value $x$ behaves as $x=2^{3/4}+\\tfrac{3}{32}\\Omega+O(\\Omega^2)$ for $\\Omega\\ll 1$ and $x=\\tfrac{3}{8}\\Omega+O(\\Omega^{-2})$ for $\\Omega\\gg 1$, showing that the unique $\\Omega=0$ saddle of the original model is on the same branch as the large-$\\Omega$ fuzzy sphere. The commuting-matrix family, parametrized by $A_3=x\\,\\sigma_3/2$ and $A_{10}=y\\,\\sigma_3/2$, has saddles at $x=\\frac{1}{4\\Omega}\\sqrt{2^{14}/3+\\Omega^4},\\,y=0$ and at $x=0,\\,y=\\frac{1}{4\\Omega}\\sqrt{2^{14}-\\Omega^4}$, both of order $1/\\Omega$, and these dominate as $\\Omega\\to 0$. The measured derivative $d\\log Z(\\Omega)/d\\Omega$ agrees with supersymmetric localization across $0.3\\le\\Omega\\le 12$ and goes as $-2/\\Omega$ as $\\Omega\\to0$, so $Z(\\Omega)\\sim\\Omega^{-2}$; the spacetime extents $\\rho_3=\\operatorname{tr}(A_a)^2$ and $\\rho_7=\\operatorname{tr}(A_I)^2$ follow the one-loop effective theory at small $\\Omega$ and show the two families exchanging dominance near $\\Omega\\sim4$.","pith_inferences":["If the two-branch structure persists at larger $N$, the polarized deformation cannot be used as a regulator that returns the original IKKT model as $\\Omega\\to0$; the commuting branch would continue to dominate no matter how smoothly the fuzzy-sphere branch connects to the undeformed saddle.","The symmetric ansatz leaves open the possibility of additional asymmetric fuzzy-sphere-type saddles; locating all stationary points of the full effective action would confirm whether the connection found here is the only one.","The one-variable analogue suggests that adding a supersymmetry-preserving mass term to the Lorentzian IKKT model could likewise generate a new dominant vacuum at infinitesimal mass, with consequences for the string-landscape interpretation.","The double-peak crossover observed at $N=2$ should sharpen as $N$ grows, since the free-energy difference between branches scales as $N^2$; a sharper transition at larger $N$ would be a testable signature."],"forward_implications":["The $\\Omega\\to0$ limit of the polarized IKKT model does not converge to the original IKKT model: the partition function diverges as $\\Omega^{-2}$ and the dominant configurations are commuting matrices with entries of order $1/\\Omega$.","The $\\Omega=0$ saddle of the original model and the fuzzy-sphere saddle at large $\\Omega$ lie on one smooth branch in the $N=2$ case, at least within the symmetric ansatz.","The change of dominance near $\\Omega\\sim4$ is a smooth crossover for $N=2$, with continuous observables and two coexisting peaks in the $\\rho_3$ and $\\rho_7$ distributions, not a sharp phase transition.","The same divergence mechanism appears in a one-variable polynomial integral, so the failure to recover the original model in the $\\Omega\\to0$ limit does not require supersymmetry or fermions.","The unpolarized extent $\\rho_7$ acts as a probe that distinguishes the two spacetime structures: it vanishes on the fuzzy-sphere branch and grows on the commuting branch."],"supporting_citations":[{"why":"defines the original IKKT matrix model, whose $\\Omega=0$ saddle is the object the paper tries to recover.","marker":"[1]"},{"why":"formulates the IIB matrix model and the commuting-matrix low-energy picture used for the commuting saddle.","marker":"[2]"},{"why":"introduces the extent-of-spacetime observable $R^2$ that the paper adapts into $\\rho_3$ and $\\rho_7$.","marker":"[10]"},{"why":"introduces the supersymmetry-preserving mass deformation that defines the polarized IKKT model.","marker":"[30]"},{"why":"provides nonperturbative matrix-model studies of fuzzy spheres, the structure that dominates at large $\\Omega$.","marker":"[34]"},{"why":"explains how D-instantons polarize into an $S^2$ worldvolume in a background flux, the physical picture of the fuzzy-sphere saddle.","marker":"[36]"},{"why":"gives the exact localization result for the partition function that the Monte Carlo data reproduce and whose $\\Omega^{-2}$ divergence the paper explains.","marker":"[38]"},{"why":"reports the localization-based phase transition at larger $N$ and supplies the comparison point for the $N=2$ crossover.","marker":"[45]"}],"fun_headline_variants":["Fuzzy sphere loses to commuting matrices at small Omega","IKKT saddle connects to fuzzy sphere, then commuting matrices dominate","Small Omega favors commuting matrices over fuzzy sphere in IKKT","Polarized IKKT: commuting matrices dominate as Omega tends to zero"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the symmetric ansatz $x_1=x_2=x_3=x$ catches the entire fuzzy-sphere saddle family and that the commuting family is the only other contributor; if another saddle family with unequal $x_a$ or nonzero $A_I$ contributes in $0.3\\le\\Omega\\le12$, the smooth-connection and completeness claims would be incomplete.","fun_headline_variants_meta":{"raw":{"variants":["Fuzzy sphere loses to commuting matrices at small Omega","IKKT saddle connects to fuzzy sphere, then commuting matrices dominate","Small Omega favors commuting matrices over fuzzy sphere in IKKT","Polarized IKKT: commuting matrices dominate as Omega tends to zero"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000784,"raw_usage":{"total_tokens":3558,"prompt_tokens":1143,"completion_tokens":2415,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":759,"completion_tokens_details":{"reasoning_tokens":2343}},"tokens_in":759,"tokens_out":2415,"duration_ms":16780,"temperature":1.0,"reasoning_tokens":2343,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T18:11:32.757366+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Evaluate the full $N=2$ saddle-point equation (8) numerically without imposing $x_1=x_2=x_3$ and without setting $A_I=0$, on a grid of $\\Omega$ in $[0.3,12]$. If any stationary configuration appears whose effective action is comparable to or lower than the two identified families, the claim that all contributing saddle points were captured fails. A direct Monte Carlo check is to start from an asymmetric initial condition near $\\Omega=1$ and see whether a third peak emerges in the $\\log\\rho_3$ or $\\log\\rho_7$ histogram.","supporting_citations":[{"cited_title":"Azuma, S","cited_arxiv_id":null,"evidence_quote":"provides nonperturbative matrix-model studies of fuzzy spheres, the structure that dominates at large $\\Omega$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"explains how D-instantons polarize into an $S^2$ worldvolume in a background flux, the physical picture of the fuzzy-sphere saddle."}],"review_version":2}