{"id":"645f38bf-72ba-4ef0-8020-9e61522c7825","arxiv_id":"2412.16987","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Type A and type B trace anomalies are shown to match between the unbroken and spontaneously broken phases of conformal field theories, with new dispersive sum rules constraining dilaton couplings to massive states.","lead":"Trace anomalies in conformal field theories resist the topological proofs used for chiral anomalies, yet this paper argues they still match between the unbroken and spontaneously broken phases, for both type A and type B. It derives dispersive sum rules constraining how the dilaton couples to the energy-momentum tensor, and verifies that a previously claimed counterexample actually obeys matching.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Matching c = c̃ rests on the unproven deep-Euclidean equality (3.4)–(3.5): broken-phase dimension -2 amplitudes are asserted to equal unbroken ones for all invariants ≫ f, without a derivation or an all-orders bound on dilaton-mediated terms.","rationale":"The reader's weakest_assumption exactly matches the load-bearing point: the equality of broken- and unbroken-phase amplitudes in the deep Euclidean limit is assumed as feature (d) and used to turn the ratio (3.4) into c = c̃. This is genuinely the hinge of the matching proof. It is a standard short-distance/OPE expectation and is supported by the explicit one-loop computations, so I do not regard the paper as wrong; but the paper labels the argument a 'general proof' while (3.5) is conditional on an unproven decoupling statement. The Appendix B check verifies the anomaly combination at one loop and at special kinematics, but does not exhibit the large-momentum asymptotic with its power corrections, nor does any argument exclude all-orders non-decoupling of the dilaton sector. The deep-IR factorization assumption used in Section 4 is a second, separate assumption affecting the sum rules, but the matching claim itself does not depend on it. Verdict remains CONDITIONAL/UNCHANGED.","tokens_in":27218,"tokens_out":13853,"duration_ms":149021,"concrete_test":"In the N=2 circular quiver model of Appendix B, perform an explicit Passarino-Veltman reduction of the triangle plus dilaton-exchange diagrams and verify, at generic external momenta q² ≫ f² with fixed ratios, that the anomaly combination -3q²B - 2q·r C + r²D tends to N with corrections no larger than O(f²/q²) and no O(1) remainder. This directly tests whether the deep-Euclidean equality used in (3.4) holds beyond the special kinematics emphasized in (B.21).","verdict_should_be":"UNCHANGED","load_bearing_attack":"Feature (d) at the start of Section 3 states that the UV structure of correlators in the broken phase is the same as in the unbroken phase. This is the only step that converts the ratio identity (3.4) into the matching conclusion c = c̃ (3.5). For (3.5) to follow, the combination Σ si(Ẽi - Ei) must vanish in the deep Euclidean limit si → ∞ at fixed ratios. The paper does not derive this from the Ward identities, nor does it show that the dilaton-exchange diagrams generated by the Goldstone coupling (3.2) decouple at large momenta; the one-loop verifications in Appendices A and B are supportive but are not an all-orders general proof. A non-decoupling term of the form f²/s_i in Ẽ_i would leave a finite f-dependent shift and invalidate (3.5). Thus the 'general proof' is conditional on an unproven decoupling assumption.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops analyticity-based sum rules for trace anomalies in four-dimensional CFTs and uses them to argue that type A and type B trace anomalies match between the unbroken phase and the spontaneously broken phase. It then derives low-energy constraints and dispersive sum rules for dilaton–energy-momentum correlators in the deep infrared limit. The matching argument in Section 3 compares the universal anomaly equations (2.8) and (3.3) in the deep Euclidean limit and concludes c = c̃; the sum rules for dilaton amplitudes in Section 4 rest on an explicitly assumed factorized deep IR structure. Appendix A gives a one-loop argument for matching, and Appendix B verifies matching in the N=2 circular quiver example previously claimed to be a counterexample.","tokens_in":27423,"tokens_out":8109,"duration_ms":71474,"significance":"If the matching result holds, it establishes a general 't Hooft-like matching for trace anomalies in broken CFTs and provides parameter-free sum rules such as (2.12), (4.23), (4.33), and (4.34). The paper also clarifies an important conceptual point: in the broken phase the type B anomaly is matched even though the relation between the anomaly and the explicit UV breaking of dilations is modified by the dilaton term in (3.7). Credit should be given for the explicit one-loop verification in Appendix B, which resolves the Niarchos–Papageorgakis–Pomoni counterexample claim, and for the systematic enumeration of the deep IR local terms in (4.22). The main limitation is that the central proof of matching depends on an unproven deep-Euclidean equality of broken-phase and unbroken-phase amplitudes, which the paper states as feature (d) rather than derives.","major_comments":[{"comment":"The matching conclusion c = c̃ follows only if the combination Σ_i s_i(Ẽ_i − E_i) vanishes in the deep Euclidean limit s_i → ∞ at fixed ratios. The paper asserts this in feature (d) and in the sentence preceding (3.5), but does not derive it from the Ward identities or from any decoupling estimate for the dilaton-mediated diagrams generated by the Goldstone coupling (3.2). A non-decoupling contribution of order f²/s_i in Ẽ_i would leave a finite f-dependent shift and invalidate (3.5). The one-loop checks in Appendices A and B are supportive evidence but not an all-orders proof. Since this step is load-bearing for the paper's central claim, the authors should either provide a proof of the deep-Euclidean equality or explicitly state it as an additional hypothesis and qualify the 'general proof' language accordingly.","section":"Section 3, feature (d) and Eqs. (3.4)–(3.5)"},{"comment":"The deep IR factorization into a normal CFT and a single dilaton sector is explicitly assumed, not derived. This assumption underlies the derivation of the sum rules in Section 4, e.g. (4.23), (4.33), and (4.34). Since these are presented as general consequences of anomaly matching, the paper should either prove the factorization or clearly state that the sum rules are conditional on this additional hypothesis. The current text acknowledges the assumption but does not indicate its scope in the abstract or introduction, so the reader may overestimate the universality of the derived constraints.","section":"End of Section 3, after Eq. (3.7)"}],"minor_comments":[{"comment":"The phrase 'general proof' is used for the matching result, but the proof in Section 3 depends on the unproven deep-Euclidean equality; please qualify the claim in the abstract and introduction.","section":"Abstract and Introduction"},{"comment":"The term '−φ̄c' inside the parentheses appears to be a typographical error; with the solution (4.36) it would generate a term linear in J, contradicting the stated O(J²) result. Please check the expression.","section":"Eq. (4.37)"},{"comment":"The table is very dense; consider adding a sentence explaining the role of the 'Schouten' column and the normalization convention for the invariant amplitudes.","section":"Table 2"},{"comment":"The distribution δ(s_j) is written without a dimensionful prefactor; clarify that this is a formal statement obtained by extrapolating to s_i = 0.","section":"Eq. (2.13)"},{"comment":"The normalization of the dilaton state |φ(0)⟩ is not specified; state the convention used so that the second term on the right-hand side is unambiguous.","section":"Eq. (3.7)"},{"comment":"The coefficients (a−1) and (c−1) are explained as subtracting the one-loop dilaton contribution, but this becomes clear only after reading the following paragraph; consider adding a parenthetical remark at first use.","section":"Section 4, after Eq. (4.22a)"}],"recommendation":"major_revision","confidential_remarks":"This is a serious contribution by established authors, and the explicit verification in Appendix B is a valuable resolution of a published counterexample claim. The central obstruction is that the 'general proof' of matching rests on an unproven deep-Euclidean decoupling assumption, which is not flagged as a hypothesis in the abstract or introduction. The paper fits the journal's scope, and the citation practice is fair; the reliance on the authors' own earlier work [5] is appropriate in context. I would encourage the editor to request that the authors either prove the decoupling or recast the paper's claims as conditional on the stated assumption."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe short version: this paper attempts a general proof of trace anomaly matching in spontaneously broken CFTs for both type A and type B, and it includes a real one-loop verification in the N=2 quiver models that were previously claimed to violate matching. The proof strategy is interesting, but the claim of a general proof overstates what is shown: the key deep-Euclidean equality is assumed, not derived.\n\nWhat is genuinely new: the sum rule formulation of anomalies via dimension −2 amplitudes; the matching argument comparing the universal anomaly equations (2.8) and (3.3) in the deep Euclidean limit; the deep IR expansion of the dilaton–e-m generating functional with the new sum rules (4.23), (4.33), (4.34); and the explicit one-loop computation in Appendix B giving c(α)=N for all α, resolving the contested counterexample of [23,24]. The observation that type B matching does not imply the same relation to explicit dilation breaking in the broken phase is also valuable.\n\nWhat the paper does well: it is careful about distinguishing unambiguous dimension −2 amplitudes from subtraction polynomials, and the one-loop check is explicit and credible. Appendix A gives a clean diagram-by-diagram argument at one loop. The paper is honest about its assumptions.\n\nThe soft spot is the step from (3.4) to (3.5). The paper asserts that in the deep Euclidean limit the broken-phase amplitudes agree with the unbroken ones because all invariants are much larger than f. That is plausible but it is not derived from the Ward identities, nor is there a bound on dilaton-mediated terms. A non-decoupling term of order f²/s_i in the Ẽi would shift the ratio and break the conclusion. The one-loop checks are supportive but not an all-orders proof. The deep IR factorization assumption in Section 4 is also an input, though it is explicitly flagged.\n\nOn balance, the central claim is likely correct, but the paper needs to either prove the decoupling or present it as a conjecture with evidence. A serious referee should focus on that step and on the analyticity assumptions behind the dispersion relations.\n\nThis is for CFT/anomaly specialists. I would send it to peer review; the sum rules and the one-loop check deserve a referee, and the authors should be asked to clarify the status of the deep-Euclidean equality.\n\nRecommended verdict: major revision with the missing derivation (or an explicit conjecture) addressed.","headline":"A serious attempt to prove trace anomaly matching with a real one-loop check, but the general proof rests on an unproven deep-Euclidean decoupling assumption.","tokens_in":27981,"tokens_out":2969,"would_cite":true,"duration_ms":26046,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81T40","81T50"],"pacs":[],"model":"deepseek-v4-flash","headline":"A general proof that type A and type B trace anomalies match in spontaneously broken CFTs.","keywords":["trace anomaly","anomaly matching","spontaneous conformal symmetry breaking","dilaton","type A and type B anomalies","dispersion sum rules","Weyl anomaly","conformal field theory"],"falsifier":"Evaluate the broken-phase dimension $-2$ amplitudes $B,C,D$ of the one-dilaton/two-energy-momentum three-point function in a model of the circular-quiver type at finite external momenta; anomaly matching predicts that the anomalous combination $-3q^2B-2q\\cdot r\\,C+r^2D$ equals the unbroken anomaly $N$ for every value of the breaking scale $v$, and any $v$- or momentum-dependent correction surviving the large-$q^2$ limit would falsify the claim.","tokens_in":26933,"feed_emoji":"🎯","tokens_out":9734,"duration_ms":77848,"temperature":0.7,"pith_summary":"This paper gives a general proof that trace anomalies are matched between a four-dimensional conformal field theory and its spontaneously broken phase, for both type A and type B anomalies. The proof uses an unambiguous characterization of trace anomalies: after the diffeomorphism Ward identities are imposed, each anomaly is encoded in a universal equation for dimension $-2$ invariant amplitudes, $\\sum_i s_i E_i = c$. The same equation holds in the broken phase, $\\sum_i s_i \\tilde E_i(s_1,s_2,s_3;f)=\\tilde c$, and because the amplitudes have the same power-like short-distance behaviour the ratio of the two equations forces $c=\\tilde c$ in the deep Euclidean limit. The paper also shows that in the broken phase the action of dilations on correlators gains a model-dependent dilaton term, so the type B anomaly is no longer tied to explicit UV breaking of dilations. The matching then controls dispersive sum rules that constrain dilaton--energy-momentum correlators and the massive spectrum.","feed_headline":"Anomaly matching proved for type A and B trace anomalies","feed_subtitle":"Equal type A and B anomaly coefficients across phases, yielding sum rules that constrain dilaton couplings.","key_machinery":"The load-bearing object is the universal anomaly equation $\\sum_{i=1}^3 s_i E_i(s_1,s_2,s_3)=c$, in which the invariant amplitudes $E_i$ have dimension $-2$ and are free of regularization ambiguity, and $c$ is the anomaly coefficient for type A or type B. Diffeo Ward identities eliminate the positive- and zero-dimension amplitudes, leaving this equation as the single expression of the anomaly. The same equation, with amplitudes depending also on the breaking scale $f$, is imposed in the broken phase. Matching follows by taking the ratio of the two equations and using the deep-Euclidean equality of the amplitudes; the discontinuities of the $E_i$ then obey the sum rule $\\int ds\\,\\mathrm{Im}_j E_j = -\\pi c$, which is the analyticity-based formulation of the anomaly. In the broken phase the dilation Ward identity is modified by the dilaton's zero-momentum coupling, which adds the model-dependent term $f\\langle \\phi(0)| T_{\\nu_1\\rho_1}\\cdots |0\\rangle$ to the transformation law.","core_discovery":"The central claim is that the spontaneously broken phase of a CFT carries the same trace anomalies as the unbroken phase: the broken-phase anomaly coefficient $\\tilde c$ in $\\sum_i s_i \\tilde E_i(s_1,s_2,s_3;f)=\\tilde c$ equals the unbroken coefficient $c$ in $\\sum_i s_i E_i(s_1,s_2,s_3)=c$. The argument compares the two universal anomaly equations in the deep Euclidean limit, where every invariant $s_i$ is much larger than the breaking scale $f$, so the broken-phase amplitudes coincide with the unbroken ones and the ratio of the equations forces $c=\\tilde c$. This is stated uniformly for type A and type B anomalies. For type B, the relation between the anomaly and the explicit violation of dilations by the UV cutoff is special to the unbroken phase; in the broken phase the dilation Ward identity acquires a nonlocal term from the massless dilaton, so the type B coefficient is no longer tied to that UV violation. The paper then derives the general structure of dilaton--energy-momentum correlators in the deep infrared and, with anomaly matching as input, obtains dispersion sum rules for the invariant amplitudes.","pith_inferences":["If the same unambiguous dimension $-2$ amplitude decomposition can be constructed in higher even dimensions, the paper's deep-Euclidean ratio argument would predict the same matching there; that extension is not carried out in this paper.","The same line of reasoning suggests that any apparent failure of type B matching in a specific model will be traceable to a missing massless or dilaton-mediated contribution, rather than to a change in the anomaly coefficient.","For massive, non-conformal flows the role of the dilaton may be played by the trace of the energy-momentum tensor or by beta-function terms; the sum rules derived here would then become constraints on the flow, as the authors note as an open question.","A testable extension would be to compute the large-momentum discontinuity of the dimension $-2$ amplitudes in a strongly coupled realization of spontaneous breaking; matching predicts it must equal the unbroken anomaly coefficient with no $f$-dependent correction."],"forward_implications":["Type A and type B anomaly coefficients are equal in every spontaneously broken phase of a given CFT, for each breaking pattern, so a single dimensionless coefficient characterizes the trace anomaly across all phases.","In the broken phase the type B anomaly coefficient is no longer fixed by the UV logarithm; the dilaton sector contributes an independent, model-dependent breaking of dilations.","The matched generating functional yields concrete dispersive sum rules, such as $\\frac{1}{\\pi}\\int \\frac{dt'}{t'^3}\\,\\mathrm{Im}\\,A_1(s=0,t')=\\frac{a-1}{2f^2}$, that constrain the couplings of the dilaton to the massive spectrum.","At one loop, each massive field matches its anomaly diagram by diagram, with the dilaton coupling fixed by $\\gamma f^2+M^2=0$.","The circular-quiver example previously presented as a type B counterexample is shown to match once the dilaton and massless-scalar contributions are included."],"supporting_citations":[{"why":"Supplies the kinematical decomposition into invariant amplitudes and the Ward identities that produce the universal anomaly equation (2.8).","marker":"[5]"},{"why":"Establishes the geometric classification of trace anomalies into type A and type B that the matching statement relies on.","marker":"[4]"},{"why":"Identifies the property that type B anomalies do not vanish for constant Weyl parameter, the feature used to separate them from dilations.","marker":"[2]"},{"why":"The claimed counterexample to type B matching on the Higgs branch that Appendix B re-examines.","marker":"[23]"},{"why":"The follow-up claim of type B mismatch whose missing massless-scalar and dilaton contributions are shown to restore matching.","marker":"[24]"},{"why":"Provides the earlier two-dilaton/two-graviton amplitude sum rule that Section 4 generalizes with the factorized deep IR structure.","marker":"[17]"},{"why":"Introduces the dilaton effective action and Wess-Zumino framework, and the four-dilaton sum rule used in the deep IR analysis.","marker":"[18]"}],"fun_headline_variants":["Broken CFTs preserve type A and B trace anomalies","Anomaly matching proven for all trace anomaly types","Type A and B trace anomalies match in broken CFTs","Broken and unbroken CFTs share trace anomaly coefficients"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof assumes feature (d) of the broken phase: at momenta large compared with the breaking scale $f$, the correlators—including their dimension $-2$ amplitudes—are the same as in the unbroken phase, so the deep Euclidean limit of the ratio of anomaly equations forces equality of the coefficients; if dilaton-mediated corrections did not decouple as the invariants go to infinity, the ratio argument would not force $\\tilde c=c$.","fun_headline_variants_meta":{"raw":{"variants":["Broken CFTs preserve type A and B trace anomalies","Anomaly matching proven for all trace anomaly types","Type A and B trace anomalies match in broken CFTs","Broken and unbroken CFTs share trace anomaly coefficients"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000326,"raw_usage":{"total_tokens":1760,"prompt_tokens":818,"completion_tokens":942,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":434,"completion_tokens_details":{"reasoning_tokens":873}},"tokens_in":434,"tokens_out":942,"duration_ms":8672,"temperature":1.0,"reasoning_tokens":873,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T05:55:40.944654+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Evaluate the broken-phase dimension $-2$ amplitudes $B,C,D$ of the one-dilaton/two-energy-momentum three-point function in a model of the circular-quiver type at finite external momenta; anomaly matching predicts that the anomalous combination $-3q^2B-2q\\cdot r\\,C+r^2D$ equals the unbroken anomaly $N$ for every value of the breaking scale $v$, and any $v$- or momentum-dependent correction surviving the large-$q^2$ limit would falsify the claim.","supporting_citations":[{"cited_title":"Comments on Trace Anomaly Matching","cited_arxiv_id":"2307.14957","evidence_quote":"Supplies the kinematical decomposition into invariant amplitudes and the Ward identities that produce the universal anomaly equation (2.8)."},{"cited_title":"Nonlocal Conformal Anom alies,","cited_arxiv_id":null,"evidence_quote":"Identifies the property that type B anomalies do not vanish for constant Weyl parameter, the feature used to separate them from dilations."}],"review_version":1}