{"id":"00e301e5-eabe-4a06-897c-6364d913cbe0","arxiv_id":"1908.10437","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"The leading large-spin anomalous dimension of double-trace operators in four-dimensional logarithmic CFTs behaves as γ0/ℓ^{τ_m}, where τ_m is the minimal twist, provided no operator has negative scaling dimension.","lead":"This paper applies the large-spin conformal bootstrap, a set of exact symmetry constraints, to logarithmic conformal field theories in four dimensions, and derives the leading large-spin correction to the dimensions of composite operators. It also checks the result against a simplified holographic model, and argues that cluster decomposition survives in these non-unitary theories when all operator dimensions are positive.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Cluster-decomposition condition is misstated: τ_m>0 requires Δ_m>ℓ_m, not merely positive dimensions; a positive-dimension spinning operator with Δ_m<ℓ_m gives a<0 and a growing anomalous dimension.","rationale":"The reader's CONDITIONAL verdict is appropriate: the bootstrap derivation is internally consistent under the explicit assumption τ_m>0, and the paper flags this assumption. However, the reader's weakest_assumption focused on the absence of operators with negative scaling dimension. The more precise load-bearing gap is that the paper's advertised condition, 'dimensions of the operators are positive,' does not guarantee τ_m>0 whenever the minimal-twist operator has nonzero spin. Since τ_m=Δ_m−ℓ_m, an operator with positive but sub-spin dimension (0<Δ_m<ℓ_m) has negative twist, which reverses the sign of the exponent a=τ_m and makes the anomalous dimension grow with spin. The abstract and conclusion therefore overstate the domain of validity of the cluster-decomposition claim. The derivation itself, with the stated τ_m>0 assumption, is not undermined; the issue is the mismatch between the formal assumption and the paper's advertised sufficient condition. A concrete analytic check, inserting a spin-2, dimension-1 minimal-twist operator into the matching equations, would confirm that positive dimensions alone are insufficient. Because the central claim is already framed as conditional, the verdict need not change, but the condition should be restated as twist positivity rather than dimension positivity.","tokens_in":21419,"tokens_out":12751,"duration_ms":148129,"concrete_test":"Re-run the power-counting step between eq. (3.4) and eq. (3.15) for a hypothetical minimal-twist operator with spin ℓ_m=2 and dimension Δ_m=1 (so τ_m=-1). Equating the v-exponents of the u^{Δφ/3} logu terms yields a=τ_m=-1, predicting γ_{0,ℓ}∼γ0ℓ, which grows with spin. This check directly settles whether the stated condition 'positive scaling dimensions' is sufficient for the cluster-decomposition conclusion; a positive-dimension spinning operator with Δ_m<ℓ_m would violate the required twist condition.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central conditional result is the large-spin behavior γ_{0,ℓ} ∼ γ0/ℓ^{τ_m} with γ0 in (3.16), and the claim that cluster decomposition holds 'as long as the dimensions of the operators are positive.' The derivation explicitly assumes τ_m>0 (Section 3, after eq. (3.5)), and the exponent a is fixed by matching the v-dependence of the logu coefficient in (3.4) and (3.15), giving a=τ_m. But for a minimal-twist operator of spin ℓ_m, the twist is τ_m=Δ_m−ℓ_m. Positivity of scaling dimensions, Δ_m>0, does not imply τ_m>0 when ℓ_m>Δ_m. In a non-unitary LogCFT there is no unitarity bound preventing, for example, a primary of dimension Δ_m=1 and spin ℓ_m=2, whose twist is −1. If such an operator is the minimal-twist exchange, the t-channel block contributes v^{τ_m/2}=v^{−1/2}, which dominates at small v, and the matching forces a=τ_m=−1<0. Then γ_{0,ℓ}∼γ0/ℓ^{-1}=γ0 ℓ, growing with spin, so the cluster-decomposition conclusion fails. The paper only rules out Δ_m<0, not Δ_m<ℓ_m. Thus the abstract's sufficient condition is too weak: the actual condition under which the derivation yields a vanishing large-spin anomalous dimension is positivity of the minimal twist, not merely positivity of scaling dimensions. This does not invalidate the algebra under the stated τ_m>0 hypothesis, but it narrows the class of LogCFTs to which the advertised conclusion applies.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper applies analytic (large-spin) conformal bootstrap to a four-dimensional logarithmic CFT built from a rank-2 logarithmic scalar multiplet. It uses the crossing equation (2.15) for the function F2 to reproduce the mean-field correlator from a sum over large-spin double-trace operators, and then derives the leading large-spin anomalous dimension of rank-3 even-spin double-trace operators, obtaining γ_{0,ℓ} ∼ γ0/ℓ^{τ_m} with γ0 given in (3.16) in terms of the minimal-twist OPE coefficients a_p, b_p, c_p. The paper also studies a higher-derivative AdS dual toy model, computes a binding energy decaying as 1/ℓ^2, and compares it qualitatively with the bootstrap result. The advertised physical conclusion is that cluster decomposition holds for LogCFTs as long as all operator dimensions are positive.","tokens_in":21765,"tokens_out":8679,"duration_ms":92174,"significance":"The conditional result is a useful extension of large-spin bootstrap techniques to non-unitary logarithmic CFTs, a direction with comparatively few higher-dimensional results. The computation is detailed, the appendices supply the main technical steps, and the authors are candid that the bulk model is toy-like and that the holographic expression (4.25) does not match (3.16). The final formula (3.16) is a crossing-symmetry constraint expressed in terms of independent OPE data rather than a closed numerical prediction, which is appropriate for a bootstrap relation. The main weakness is that the advertised sufficient condition for cluster decomposition is stated too broadly: positivity of scaling dimensions does not imply positivity of the minimal twist that controls the large-spin behavior.","major_comments":[{"comment":"The sufficient condition for the cluster-decomposition conclusion is misstated. The derivation fixes the exponent a by comparing the v-dependence of (3.4) and (3.15), giving a = τ_m, and the decay γ_{0,ℓ} ∼ γ0/ℓ^{τ_m} requires τ_m > 0. The paper explicitly assumes 'the operator dimensions are always positive and τ_m > 0' just before (3.5), but the Abstract and the concluding paragraphs replace this with the weaker statement 'as long as the dimensions of the operators are positive.' For a spinning minimal-twist operator, τ_m = Δ_m − ℓ_m, so Δ_m > 0 does not imply τ_m > 0; for example, in d = 4 a primary with Δ_m = 1 and ℓ_m = 2 has τ_m = −1. If such an operator is the minimal-twist exchange, the t-channel block behaves as v^{τ_m/2} = v^{-1/2}, the matching forces a = −1, and γ_{0,ℓ} ∼ γ0 ℓ grows with spin, so the advertised cluster-decomposition conclusion fails. The correct condition under which the derivation yields a vanishing large-spin anomalous dimension is positivity of the minimal twist (equivalently Δ_m > ℓ_m for the minimal-twist operator), not merely positivity of scaling dimensions. The Abstract and Conclusion should be amended to state this narrower condition, or the class of LogCFTs considered should be restricted accordingly.","section":"Section 3, Eq. (3.16); Abstract and Conclusion"}],"minor_comments":[{"comment":"Because (4.25) does not match (3.16) and the bulk derivation relies on the ad hoc prescription (B.13) for Γ(1+q) at negative integer q, the holographic discussion should be framed as a heuristic check of the 1/ℓ^2 parametric behavior rather than as a derivation or quantitative confirmation of the bootstrap result.","section":"Section 4, Eq. (4.25)"},{"comment":"The lower limit ℓ0 in the integrals in (2.30) is not defined; a sentence identifying it as a large-spin cutoff (for instance, ℓ0 ≫ 1) would avoid ambiguity.","section":"Section 2, Eq. (2.30)"},{"comment":"The passage from the general derivative structure of the logarithmic OPE to the 'relevant' terms in (3.12) is quite compressed; a short explanation of why the discarded terms cannot contribute to the coefficient of log u would improve readability.","section":"Section 3, Eq. (3.12)"},{"comment":"Reference [76] lists the third author as 'F. Lalo'; the standard spelling is 'Laloë'.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The main issue is a mismatch between the derivation's explicit assumption τ_m > 0 and the broader claim in the Abstract/Conclusion that positive scaling dimensions suffice for cluster decomposition. This is a load-bearing statement but is easily fixable by editing the wording and explicitly restricting the class of LogCFTs. I do not see a fatal flaw in the bootstrap algebra under the stated τ_m > 0 hypothesis; the bulk section is clearly a toy and should not carry quantitative weight."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Read Banerjee-Dey on analytic bootstrap for LogCFT. The core calculation is plausible and worth a serious referee, but the abstract's cluster-decomposition claim is too strong: the condition actually used is τ_m > 0, not merely positive scaling dimensions. For a spinning operator τ_m = Δ_m − ℓ_m, so Δ_m > 0 does not rule out a negative twist (e.g., Δ_m=1, ℓ_m=2). If such an operator were the minimal-twist exchange, a = τ_m < 0 and γ0,ℓ would grow with spin, violating cluster decomposition. The authors do state the τ_m>0 assumption in Section 3, so the algebra is consistent; the problem is that the abstract and conclusion replace it with a weaker condition that does not guarantee the advertised result. This is a precise, correctable issue.\n\nWhat is actually new: this is the first application of the large-spin analytic bootstrap to rank-2 logarithmic multiplets in d=4, as far as I know. The mean-field consistency check in Section 2 is worked out carefully, and Eq. (3.16) for the leading large-spin anomalous dimension of rank-3 even-spin double-trace operators is a new result, even though it is expressed in terms of undetermined OPE coefficients. The authors are honest that γ0 is not a numerical prediction.\n\nSoft spots in proportion: the bulk comparison in Section 4 is qualitative and the authors admit Eq. (4.25) does not match Eq. (3.16); that is acceptable for a toy model but provides no independent check of the coefficient. The t-channel argument assumes a single minimal-twist operator dominates, and the OPE data (a_p,b_p,c_p) are not computed, so the formula is a constraint rather than a prediction. These are acknowledged limitations, not hidden flaws. Exposition is a bit rough in places, but the steps are checkable.\n\nWho this is for: people working on LogCFTs or on analytic bootstrap in non-unitary theories. It deserves peer review—a conditional accept with requests to correct the cluster-decomposition statement and to clarify that the result applies only when τ_m > 0. I would cite it if I worked on LogCFT bootstrap.","headline":"A plausible first application of the large-spin analytic bootstrap to LogCFTs, but the abstract's cluster-decomposition claim is too strong: the derivation requires positive minimal twist τ_m, not merely positive scaling dimensions.","tokens_in":22307,"tokens_out":3786,"would_cite":true,"duration_ms":35507,"reading_group":"maybe","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 shows that crossing symmetry fixes the leading large-spin anomalous dimension of even-spin rank-3 double-trace operators in a logarithmic CFT, so cluster decomposition holds whenever all operator dimensions are positive.","keywords":["logarithmic conformal field theory","analytic bootstrap","large-spin limit","anomalous dimension","cluster decomposition","AdS/CFT correspondence","non-unitary CFT","conformal crossing symmetry"],"falsifier":"Take a concrete solvable four-dimensional LogCFT whose spectrum has all positive dimensions, such as a perturbed logarithmic generalized free field, compute its four-point function to the order that isolates the $\\log u$ coefficient, and compare the large-spin anomalous dimension with $\\gamma_0/\\ell^{\\tau_m}$ using (3.16); a mismatch in the exponent or a divergent prefactor would disprove the claim.","tokens_in":21166,"feed_emoji":"","tokens_out":18045,"duration_ms":155457,"temperature":0.7,"pith_summary":"Logarithmic conformal field theories (LogCFTs) are CFTs whose correlation functions contain logarithms because the dilatation operator acts non-diagonally; they appear in settings from disordered systems to string theory. Standard numerical bootstrap methods require unitarity, which LogCFTs lack, so this paper applies the analytic large-spin bootstrap instead. It studies the four-point function of a rank-2 multiplet of logarithmic scalars and shows that crossing symmetry, through the bootstrap equation (2.15), determines the leading anomalous dimension of the even-spin rank-3 double-trace operators to be $\\gamma_{0,\\ell} \\sim \\gamma_0/\\ell^{\\tau_m}$, where $\\tau_m$ is the twist of the minimal-twist operator and $\\gamma_0$ is given in (3.16) in terms of that operator's OPE data. Consequently, if all operator dimensions are positive, the anomalous dimension vanishes at infinite spin and cluster decomposition holds for these non-unitary theories. A holographic check with $\\tau_m=2$ reproduces the parametric $1/\\ell^2$ decay as the binding energy of two rapidly orbiting particles in AdS.","feed_headline":"Logarithmic CFTs keep cluster decomposition at large spin","feed_subtitle":"Crossing symmetry pins the anomalous dimension of double-trace operators, so non-unitary LogCFTs decouple at infinite spin.","key_machinery":"The load-bearing mechanism is the large-spin analytic bootstrap, applied to the crossing-symmetric bootstrap equation (2.15). In the limit $v \\ll u \\ll 1$ with $\\ell \\gg 1$ and $v\\ell^2$ held fixed, the conformal blocks reduce to modified Bessel functions $K_0(2\\ell\\sqrt{v})$, and comparing the $\\log u$ coefficient on both sides of the equation forces the anomalous-dimension exponent to equal the minimal twist $\\tau_m$ (the twist is dimension minus spin). The rank-3 logarithmic multiplet, the set of three even-spin double-trace operators $S_1, S_2, S_3$ built from two logarithmic scalars, enters through the OPE data $a_p, b_p, c_p$ packaged in the coefficient $D_{\\mathcal{O}}$ of (2.18), which fixes $\\gamma_0$ in (3.16). The assumption that all operator dimensions are positive enters here: $\\tau_m > 0$ makes the exponent positive and the anomalous dimension vanish at infinite spin.","core_discovery":"The central claim is that the crossing equation (2.15) for the function $F_2(u,v)$, which encodes the four-point function of two rank-2 logarithmic scalars, determines the leading large-spin behaviour of the double-trace operators that appear in their operator product expansion (OPE). Matching the coefficient of $\\log u$ between the s-channel expansion, dominated by large-spin double-trace operators, and the t-channel exchange of the minimal-twist operator fixes the exponent in the ansatz $\\gamma_{0,\\ell} \\sim \\gamma_0/\\ell^a$ to be $a = \\tau_m$. The prefactor $\\gamma_0$ is given by (3.16) and depends on the OPE coefficients $a_p, b_p, c_p$ of the rank-3 logarithmic multiplet through the block coefficients (2.18). The paper further argues that, because $\\gamma_{0,\\ell} \\to 0$ as $\\ell \\to \\infty$ whenever $\\tau_m > 0$, the cluster decomposition principle survives in this class of non-unitary LogCFTs, and it verifies the parametric decay in a simplified holographic model where the minimal twist is 2.","pith_inferences":["Editorial extension: the same crossing equation should also determine subleading $1/\\ell$ corrections to $\\gamma_{0,\\ell}$; computing them in the integral-transform formulation of the bootstrap would give a sharper test of the $\\tau_m > 0$ assumption.","Editorial extension: if a LogCFT with a negative-dimension operator were found, the exponent $a = \\tau_m$ would become negative and the anomalous dimension would grow with spin, possibly marking a breakdown of cluster decomposition in physically realised LogCFTs.","Editorial extension: because the derivation does not use positivity of OPE coefficients, the large-spin bootstrap could be applied to other non-unitary CFTs, where the same $\\log u$ matching would predict analogous large-spin scaling.","Editorial extension: the appearance of dimension-derivatives of conformal blocks in (3.15) suggests that logarithmic partners contribute through derivative-type OPE data; testing whether higher-rank logarithmic multiplets obey the same pattern would extend the result beyond rank 2."],"forward_implications":["Any LogCFT with positive operator dimensions has a large-spin sector whose double-trace anomalous dimensions vanish as $\\ell^{-\\tau_m}$, so the operators become asymptotically free at infinite spin.","Cluster decomposition, usually proven using unitarity, holds for this class of non-unitary LogCFTs in four dimensions, and the bootstrap argument is expected to extend to general spacetime dimension.","The leading large-spin spectrum is universal: crossing symmetry forces every LogCFT to contain a tower of double-trace operators with twists $2\\tau + n$, independent of the details of the theory.","In the holographic dual, the anomalous dimension equals the binding energy of two rapidly rotating scalars in AdS; the paper's $\\tau_m = 2$ calculation gives a $1/\\ell^2$ energy shift, matching the bootstrap prediction parametrically.","The method provides a route to bootstrap non-unitary theories where numerical approaches fail because OPE coefficients are not positive."],"supporting_citations":[{"why":"Defines the four-dimensional LogCFT of a rank-2 logarithmic scalar multiplet, gives the four-point function and its decomposition into conformally invariant functions, and supplies the OPE structure used throughout.","marker":"[30]"},{"why":"Supplies the large-spin analytic bootstrap method and the conformal-block expansions in the kinematic limit used in the paper's crossing equation.","marker":"[46]"},{"why":"Furnishes the holographic interpretation of anomalous dimensions as binding energies of two rapidly rotating particles in AdS, used for the paper's bulk check.","marker":"[48]"},{"why":"Gives the mean-field OPE coefficients for double-trace operators, whose large-spin asymptotics feed the s-channel sum.","marker":"[68]"},{"why":"Provides the identity that equates derivatives of conformal blocks with respect to external dimensions, used to show the odd-spin contribution vanishes for identical scalars.","marker":"[65–67]"},{"why":"Supplies the large-spin integral representations and Bessel-function asymptotics that turn the spin sum into the power-law behaviour used in the bootstrap analysis.","marker":"[69]"},{"why":"Provides analogous large-spin correlator techniques used in the integrals for the anomalous-dimension extraction.","marker":"[70]"},{"why":"Provides the integral representation of the hypergeometric function used to isolate logarithm terms in the t-channel conformal blocks.","marker":"[71]"}],"fun_headline_variants":["LogCFTs respect cluster decomposition at large spin","Large spin bootstrap tames logarithmic CFTs","LogCFT anomalous dimensions pinned by crossing at large spin","Non-unitary LogCFTs still cluster decompose at infinite spin","Bootstrap shows LogCFTs obey cluster decomposition"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes every operator in the theory has positive scaling dimension; if an operator with negative dimension existed, the twist (dimension minus spin) that controls the crossed channel would be negative and the anomalous dimension would grow rather than vanish at large spin.","fun_headline_variants_meta":{"raw":{"variants":["LogCFTs respect cluster decomposition at large spin","Large spin bootstrap tames logarithmic CFTs","LogCFT anomalous dimensions pinned by crossing at large spin","Non-unitary LogCFTs still cluster decompose at infinite spin","Bootstrap shows LogCFTs obey cluster decomposition"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000163,"raw_usage":{"total_tokens":1236,"prompt_tokens":929,"completion_tokens":307,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":545,"completion_tokens_details":{"reasoning_tokens":231}},"tokens_in":545,"tokens_out":307,"duration_ms":3352,"temperature":1.0,"reasoning_tokens":231,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T10:43:47.038140+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a concrete solvable four-dimensional LogCFT whose spectrum has all positive dimensions, such as a perturbed logarithmic generalized free field, compute its four-point function to the order that isolates the $\\log u$ coefficient, and compare the large-spin anomalous dimension with $\\gamma_0/\\ell^{\\tau_m}$ using (3.16); a mismatch in the exponent or a divergent prefactor would disprove the claim.","supporting_citations":[{"cited_title":"http://dlmf.nist.gov/","cited_arxiv_id":null,"evidence_quote":"Provides the integral representation of the hypergeometric function used to isolate logarithm terms in the t-channel conformal blocks."}],"review_version":1}