{"id":"b5887020-8d1c-4c26-947f-f6e4c63ec043","arxiv_id":"2412.10499","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Melonic large-N CFTs are exactly the conformal mean field theories that extremize the universal part of the sphere free energy under linear IR marginality constraints.","lead":"The paper finds that the conformal data of large-N melonic CFTs, including SYK models, tensor models, and vector models, are determined by extremizing the universal sphere free energy of generalized free fields subject to linear marginality constraints. The result extends the F and a-maximization principles of supersymmetric theories to non-supersymmetric solvable CFTs in continuous dimension.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The proof in §5.4 discards R-dependent extrema of the conformal slice, yet the completeness of the melonic classification depends on showing no valid flat-space IR CFT is missed; the current argument appears to assume the conclusion.","rationale":"The reader's verdict CONDITIONAL is appropriate and aligned with the identified weakness. The paper proves that the actual IR solution extremizes the constrained tilde F, but the abstract claims a complete classification. The proof of the converse—that all constrained extrema correspond to actual IR solutions of the full theory—is not given in detail, and the paper's own §8 questions about missing solutions at integer d and the choice of vacuum indicate this gap. My concrete test would directly probe completeness by examining the conformal solutions of the full SDE on the sphere, which is the natural setting where the discarded R-dependent extrema could either appear or be ruled out. This does not change the verdict but sharpens the conditional: the classification should be stated as a conjecture or the completeness proof should be supplied.","tokens_in":27490,"tokens_out":1336,"duration_ms":11940,"concrete_test":"Directly analyze the conformal solutions of the full 2PI Schwinger-Dyson equations (5.15) on the sphere for a multi-field melonic model (e.g., the quartic Yukawa model of §7.1) at generic non-integer d, without first imposing the flat-space ansatz mm = 0. If every conformal solution of (5.15) that is R-independent as R → ∞ has mm = 0 and reproduces the tilde F-extremization results, completeness holds; if any R-dependent conformal solution survives the limit with different scaling dimensions, the classification is incomplete.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim is complete classification: every melonic CFT is a constrained extremum of tilde F. The proof in §5.3–5.4 only shows that the actual quantum solution, evaluated on the conformal slice, extremizes tilde F({Delta, g}) as a function. The converse direction is the completeness claim: given an extremum of tilde F satisfying mm = 0, one can construct a solution of the full 2PI equations and hence a valid IR CFT. Section 5.4 asserts only that R-independent extrema give Gs that extremize Gamma[G]|Sd; it does not show that all conformal solutions of the full Schwinger-Dyson equations are captured by the constrained tilde F-extremization. Since the full SDE (5.15) is an integral equation whose conformal solutions could in principle have R-dependent normalizations at finite R, the classification might miss valid IR CFTs if the sphere-to-flat-space limit does not commute with the large-N/melonic resummation. The paper's own discussion (§8 item 4) concedes gaps at certain integer d, reinforcing that completeness is asserted rather than proven.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper claims that the IR conformal data of all 'melonic' large-N CFTs (generalized SYK models, tensor models, and certain vector models) are determined by a constrained extremization of the universal part of the sphere free energy, tilde F, computed for generalized free fields with trial scaling dimensions. The constraints are linear marginality conditions sum_phi q^m_phi Delta_phi = d for each melonic interaction. The derivation is given twice: first via the 2PI effective action on the sphere (Section 5), and then via direct Schwinger-Dyson resummation in the conformal limit (Appendix A). The paper applies the procedure to the quartic Yukawa model, the Popovic model, a supersymmetric component model, and the large-n vector models, where it reproduces the standard 1/n anomalous dimension. It also discusses additional extrema, complex solutions, and gaps at integer dimensions.","tokens_in":27856,"tokens_out":15097,"duration_ms":152695,"significance":"If the central claim is fully established, the paper provides a genuinely unifying principle for a large class of exactly solvable large-N CFTs and extends the logic of F- and a-maximization to non-supersymmetric theories in continuous dimension. The 2PI derivation is structurally sound and the reproduction of the known vector-model anomalous dimension (6.6) is a valuable check. The paper also deserves credit for giving two distinct routes to the result, for tabulating concrete numerical solution spaces, and for being honest about unresolved issues such as missing solutions at integer d. The main weakness is that the completeness direction of the classification is asserted rather than rigorously proved, and the status of spurious extrema of the constrained tilde F problem is not fully clarified.","major_comments":[{"comment":"The completeness direction of the classification is asserted rather than proved. The text states that R-dependent extrema of tilde F 'do not satisfy (5.15)', but it does not demonstrate this, and it immediately discards 'other solutions' to (5.23a). Moreover, (5.23a) also admits solutions with tilde M(m_m)=0 for positive integer m_m or with g_m=0, and these are not analysed as separate cases. Since the abstract claims a 'complete classification', the proof should supply the missing converse: for a nonzero melon, a conformal propagator solves the flat-space Schwinger-Dyson equation (5.15) only when m_m=0, and then the coupling is fixed by (A.17c). Appendix A.3 already contains the dimensional-analysis argument that achieves this, but the main text does not connect it to the R-dependence discussion in Section 5.4. Please either promote that argument into the main proof or qualify the 'complete classification' claim.","section":"Section 5.4, Eqs. (5.23)-(5.24) and Appendix A"},{"comment":"The variational problem treats the Lagrange multipliers g'_m as unconstrained numbers, but for a physical melonic theory they are proportional to positive squared couplings in the conventions of (5.26). Equation (A.17c) then imposes a sign condition on the derivatives d tilde F_phi/dDelta_phi that is never checked. If the word 'precisely' in the abstract is intended as an iff statement, the paper should either prove that every IR-wedge extremum with m_m=0 is realizable with physical couplings, or explicitly state that the classification is one-way: every melonic CFT arises this way, but not every extremum of tilde F necessarily corresponds to a physical QFT vacuum. This matters because several solutions displayed in Figures 6-8 are complex or lie outside the IR wedge, and no criterion is given for selecting the physical ones.","section":"Sections 4.1, 5.4 and 7, Eqs. (4.2), (5.25), (A.17c)"},{"comment":"The derivation of the extremization equations is obscured by an apparent power-of-R inconsistency. Equation (5.18) contains (2R)^{2m_m}, while the subsequent extremization equations (5.23a,b) and the expansion (5.24) use (2R)^{-2m_m}. With the definition of tilde M in (5.21), the extremization equations do not follow from (5.18) as written. The sign of the exponent in (5.18) should be corrected, or the definition of tilde M adjusted, so that the reader can verify the variational steps explicitly.","section":"Section 5.3, Eq. (5.18)"}],"minor_comments":[{"comment":"Reference [59] is cited as 'Kutsakov' in Section 4; the correct spelling is 'Kutasov'.","section":"References"},{"comment":"The abstract says 'the knownF and a-maximization procedures'; an article and spacing are missing ('the known F- and a-maximization procedures').","section":"Abstract and Section 3.2"},{"comment":"The sentence defining the quartic Yukawa model as 'hλprismatic' appears garbled; the notation should be defined or removed.","section":"Section 7.3, Eq. (7.6)"},{"comment":"The role of N versus M^q in the tensor-model normalization is explained in the text, but it would help to state explicitly in (5.13) that the sum over fields counts each dynamical field once and that N is the total number of colors entering the large-N counting.","section":"Section 5.2.1, Eq. (5.11)"}],"recommendation":"major_revision","confidential_remarks":"The paper is a good match for JHEP and the central physical idea is attractive. The main issue is not the forward direction (quantum solution implies extremum of tilde F) but the completeness and realizability of the converse, which should be fixed before publication. No concerns about prior work credit; the manuscript cites the relevant SYK, tensor-model, and vector-model literature appropriately."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know two things. First, this paper largely delivers the goods: it proves a universal principle for the IR of melonic large-N CFTs — extremize the sphere free energy of generalized free fields subject to linear marginality constraints — and it does so via the 2PI effective action, which is the right tool. Second, the 'complete classification' claim in the abstract is a bit stronger than what the proof actually establishes. The gap is narrow, not a fatal flaw.\n\nWhat is genuinely new is the 2PI derivation in Section 5. Earlier work solved individual models (SYK, tensor, vector) one by one; here the authors show the mechanism is universal and correctly reproduces known results, including the large-n vector-model anomalous dimension (6.6). The Lagrange-multiplier interpretation of the running couplings is elegant, and the appendix gives an independent diagrammatic route that corroborates the main argument. I also appreciate the careful discussion of the IR wedge and the treatment of couplings that run to zero. The large-n vector models as a limiting case is a useful pedagogical bridge and explains the multiplicity of vacua in terms of poles of the Plancherel measure.\n\nThe main soft spot is the completeness direction. Section 5.4 asserts that only R-independent extrema of the conformal slice give extrema of the full 2PI functional; R-dependent extrema are discarded as inconsistent with an IR fixed point. That is physically plausible, but it is asserted rather than proven that no R-dependent solution could survive a different limiting procedure. Moreover, the paper itself concedes that at certain integer dimensions no real solutions exist (Section 8, item 4), so 'complete classification' is aspirational. I don't see circularity here — the marginality constraints emerge from the conformal limit, not from assuming the answer — but the converse direction (all conformal SDE solutions are captured) is implicit and deserves one or two clarifying sentences.\n\nThe mathematics is otherwise coherent. The citation pattern is appropriate, with due credit to the model-by-model predecessors. The paper is for anyone working on SYK-like models, tensor field theories, or large-N vector CFTs. It deserves a serious referee; I would send it out. My own verdict: the central result holds, and the abstract should be toned down from 'complete classification' to something like 'classification of the conformal vacua within the stated assumptions.'","headline":"The constrained tilde-F extremization principle for melonic CFTs is real and largely proved, but the 'complete classification' wording outruns the proof.","tokens_in":28245,"tokens_out":3876,"would_cite":true,"duration_ms":37658,"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":"A universal variational principle fixes the conformal data of all melonic large-N CFTs: extremize the universal part of the sphere free energy of the corresponding generalized free field theory, subject only to the infrared marginality of…","keywords":["melonic CFTs","sphere free energy","F-maximization","large-N limit","2PI effective action","SYK model","tensor field theories","generalized free fields"],"falsifier":"Compute the full two-point Schwinger-Dyson solution of a multi-field melonic model at a non-integer dimension, say the quartic Yukawa model at $d=2.5$, and check whether every solution lying in the infrared wedge also satisfies $m_m = 0$ for every melon and extremizes $\\widetilde{F}$; a solution with $m_m \\neq 0$ that still gives a consistent flat-space CFT would refute the classification.","tokens_in":27284,"feed_emoji":"🌀","tokens_out":10565,"duration_ms":85685,"temperature":0.7,"pith_summary":"The paper claims that every melonic large-$N$ CFT -- including the generalized SYK models, tensor field theories, and vector models -- is fully determined by a single constrained extremization principle. One extremizes $\\widetilde{F}$, the universal part of the sphere free energy of a generalized free field theory with the same field content but arbitrary conformal dimensions, subject only to the linear constraints that each melonic interaction is marginal in the infrared. If true, this gives a complete classification of melonic CFTs and extends the $F$- and $a$-maximization methods known from supersymmetric theories to non-supersymmetric CFTs in continuous dimension $d$. The proof runs through the two-particle-irreducible effective action and, equivalently, through Schwinger-Dyson resummation, showing that the running couplings become Lagrange multipliers that enforce the marginality constraints. A sympathetic reader would care because it reduces a zoo of individually solved models to one variational principle and connects extremization of free energy with counting the number of infrared degrees of freedom.","feed_headline":"One extremization rule determines every melonic CFT","feed_subtitle":"Sphere free energy plus an infrared marginality condition reproduces SYK, tensor, and vector model fixed points.","key_machinery":"The load-bearing object is $\\widetilde{F}$, the universal part of the sphere free energy, defined by $\\widetilde{F} = -\\sin(\\pi d/2) \\log Z_{S^d}$ for a collection of generalized free fields; it is finite in continuous dimension and interpolates between the Weyl anomaly coefficients and the odd-dimensional sphere free energy. The proof passes through the 2PI effective action $\\Gamma[\\{G_\\phi\\}]$ evaluated on the sphere with conformal propagators: the exact quantum solution extremizes this functional, and in the melonic limit the interaction vertices collapse into linear constraints on the scaling dimensions, with the renormalized squared couplings acting as Lagrange multipliers. The key technical fact is that the complete melon integral $\\widetilde{M}(m_m)$ vanishes linearly in the marginality parameter $m_m = \\sum_\\phi q^m_\\phi \\Delta_\\phi - d$ near $m_m = 0$, which is why only the infrared marginality constraints survive in the conformal limit.","core_discovery":"The central claim is that, for any melonic QFT in $d$ dimensions, the infrared CFT is specified by constrained extremization of $\\widetilde{F}$, defined for a mean field theory with the same field content but arbitrary trial scaling dimensions $\\Delta_\\phi$. Concretely, with melonic interactions of schematic form $g_m \\prod_\\phi \\phi^{q^m_\\phi}$, the physical dimensions extremize $\\widetilde{F}(\\{\\Delta_\\phi\\}) = \\sum_\\phi \\widetilde{F}_\\phi(\\Delta_\\phi, \\rho'_\\phi)$ subject to $\\sum_\\phi q^m_\\phi \\Delta_\\phi - d = 0$ for every interaction that does not run to zero. Equivalently, the melonic CFTs are precisely the conformal mean field theories -- theories whose correlators are sums of products of two-point functions -- with constrained extremal $\\widetilde{F}$. The paper establishes this by showing that the conformal slice of the 2PI effective action on the sphere reduces exactly to $\\widetilde{F}$ plus linear constraints, and verifies the procedure on the quartic Yukawa, Popović, supersymmetric, and large-$n$ vector model examples.","pith_inferences":["If the claim is correct, the same constrained extremization may apply beyond strictly melonic diagrams to any large-$N$ limit dominated by factorization; the test would be whether non-melonic corrections shift the extremum at order $1/N$.","The coincidence between melonic and supersymmetric extremization suggests that, in unitary integer dimensions, the physical vacuum might always be a maximum of $\\widetilde{F}$; if so, the vacuum selection problem among the discrete infrared solutions reduces to maximizing the count of infrared degrees of freedom.","Tuning the free propagator changes the infrared wedge, so the same procedure should reproduce long-range and $\\Box^k$ CFTs; a direct check would be to run the extremization for the long-range SYK line and compare with its known solution.","Once the extremal dimensions are known, mean field theory fixes all higher-point functions and OPE data, so the newly predicted vacua come with concrete spectral predictions that could be tested by direct diagrammatic computation."],"forward_implications":["Every melonic CFT -- SYK-like, tensor, or vector -- can be solved by one constrained extremization instead of a model-by-model Schwinger-Dyson analysis.","The procedure extends $F$- and $a$-maximization to non-supersymmetric CFTs in continuous dimension, so supersymmetric extremization results can be imported directly for melonic SCFTs.","The classification is complete in the strict large-$N$ limit: the infrared CFT is exactly a mean field theory with constrained extremal $\\widetilde{F}$, with all finite-symmetry details reduced to the dimensions of the symmetry representations.","The extremization generically produces a discrete set of candidate vacua, including complex scaling dimensions at some values of $d$ and missing solutions at certain integer dimensions, which can be compared against direct model computations.","Because $\\widetilde{F}$ interpolates between $a$ and $F$, the result supports interpreting the infrared fixed point as extremizing the effective number of infrared degrees of freedom, in the spirit of the generalized $c$, $F$, and $a$ theorems."],"supporting_citations":[{"why":"supplies the definition of the universal part of the sphere free energy and the $F$/$a$-maximization analogy the paper generalizes.","marker":"[1]"},{"why":"provides the $F$-theorem and $F$-maximization background and the constraint structure used for the supersymmetric comparison.","marker":"[2]"},{"why":"supplies the 2PI effective action formalism for SYK and tensor models, which is the central proof tool.","marker":"[37]"},{"why":"provides the generalized-free-field $\\widetilde{F}$ results and the shadow/inverse propagator identities used in the extremization.","marker":"[51]"},{"why":"defines the generalized SYK models, one of the principal classes of melonic CFTs the claim covers.","marker":"[4]"},{"why":"gives the large-$n$ vector model critical exponents used as the benchmark pattern for the extremization solutions.","marker":"[7]"},{"why":"supports the generalized $F$-theorem and the interpolation between $a$ and $F$ that motivates counting infrared degrees of freedom.","marker":"[34]"},{"why":"contains the quartic Yukawa melonic model and the general features of melonic CFTs used as worked examples.","marker":"[27]"}],"fun_headline_variants":["Extremize F to get every melonic CFT","Constrained F-extremization classifies melonic CFTs","One extremization rule for SYK, tensor, and vector CFTs","Sphere free energy extremization fixes all melonic CFTs","Melonic CFTs from a single F-extremization principle"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes that the only infrared solutions that can be conformally mapped to flat space are the $R$-independent extrema of $\\widetilde{F}$; if some $R$-dependent extremum of the conformal slice nevertheless produced a valid flat-space CFT through a different limiting procedure, the proposed classification would miss it.","fun_headline_variants_meta":{"raw":{"variants":["Extremize F to get every melonic CFT","Constrained F-extremization classifies melonic CFTs","One extremization rule for SYK, tensor, and vector CFTs","Sphere free energy extremization fixes all melonic CFTs","Melonic CFTs from a single F-extremization principle"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000944,"raw_usage":{"total_tokens":4063,"prompt_tokens":1005,"completion_tokens":3058,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":621,"completion_tokens_details":{"reasoning_tokens":2967}},"tokens_in":621,"tokens_out":3058,"duration_ms":618525,"temperature":1.0,"reasoning_tokens":2967,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T15:53:59.883031+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the full two-point Schwinger-Dyson solution of a multi-field melonic model at a non-integer dimension, say the quartic Yukawa model at $d=2.5$, and check whether every solution lying in the infrared wedge also satisfies $m_m = 0$ for every melon and extremizes $\\widetilde{F}$; a solution with $m_m \\neq 0$ that still gives a consistent flat-space CFT would refute the classification.","supporting_citations":[{"cited_title":"Zinn-Justin,Quantum Field Theory and Critical Phenomena","cited_arxiv_id":null,"evidence_quote":"gives the large-$n$ vector model critical exponents used as the benchmark pattern for the extremization solutions."}],"review_version":1}