{"id":"161d8032-a4d5-4e47-8154-5ac7bef13747","arxiv_id":"2504.14404","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"The ERG flow of the USp(2N) singlet sector of a free 3D U(2N) scalar theory yields a bulk AdS4 cubic action that is on-shell equivalent to Yang-Mills plus a field-strength-cubed term.","lead":"This paper shows how the cubic interaction of a force field in four-dimensional anti-de Sitter space can be derived from the renormalization group flow of a simple three-dimensional theory of free particles. If correct, it supports the idea that the forces of a holographic bulk theory emerge from boundary data rather than being put in by hand.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The cubic vertex rests on a simplified low-energy propagator whose O(p/Λ_b) error is asserted, not derived; this does not control the relative coefficients in Eq. (44) that fix g' = ig/6.","rationale":"I agree with the reader's weakest_assumption: the simplified-propagator substitution in Section 5.2 is the single most load-bearing step. The entire cubic vertex, including the relative coefficients that determine g'/g, is obtained by replacing the complicated ∆_l with a tractable one. The error bound is asserted, and it is a correlator-level statement, not a statement about the tensor coefficients of the action. Because the z^2 and (1+rz) terms in Eq. (44) have different weights in the z-integral after on-shell substitution, a small error in the propagator can translate into an O(1) change in g'/g. For the central claim—that the bulk action is the gauge-fixed form of a local gauge-invariant action with a specific F^3/YM ratio—this is precisely the step that must be controlled. The reader's other concern (on-shell-only matching) is less load-bearing because holographic boundary correlators are computed via the classical on-shell action, so an on-shell equivalence is the physically relevant one; and the locality concern is explicitly acknowledged by the authors as an open issue. The paper does have independent support: the final action reproduces the conformal Ward identity form of [47], and the construction by construction reproduces boundary correlators if the ERG equation is solved exactly. But the simplified propagator breaks that exactness. Hence the verdict should remain CONDITIONAL: the claim is plausible and well-structured, but it hinges on an approximation whose effect on the key ratio is unexamined. My concrete test—comparing the bulk-derived 3-point function with the exact free-field triangle diagram—would settle whether the approximation preserves the physically relevant coefficients.","tokens_in":27485,"tokens_out":10121,"duration_ms":93992,"concrete_test":"Compute the leading-order 1/N three-point function of the SU(2) currents in the free USp(2N) vector model from the explicit scalar triangle diagram and compare it with Eq. (143) using g = 1/(√N 128πγ^3) and g' = ig/6. If the ratio of the qµrνpρ coefficient to the δµνpρ coefficient, and the overall normalization, do not match the free-field result, the simplified-propagator substitution in Section 5.2 has altered the relative weights of the Yang-Mills and F^3 terms and the central claim fails. This test is decisive because the free-field correlator is known exactly; it does not rely on the bulk action's own approximations.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim—that the ERG-derived cubic term of the bulk spin-1 action matches, on-shell, the cubic vertex of Yang-Mills plus F^3 with g' = ig/6 (Section 6, Eqs. (46)-(60))—depends entirely on the tensor structure of the k-integral in Eq. (44). That integral is not evaluated with the actual low-energy ERG propagator ∆_l forced by the regulator constraints of [32]; instead, Section 5.2 replaces it by a 'simple' propagator, stating that the error in correlators is O(p/Λ_b). This error estimate is asserted, not derived, and it is about correlators, not about the coefficients of the action. The terms in Eq. (44) scale differently in z: the z^2 qµ rν pρ term and the δµν pρ(1+rz) terms. Their relative coefficient determines the ratio g'/g after the on-shell replacement ∂_z a = -p a and the surface-term manipulation in Eqs. (57)-(60). An O(p/Λ_b) error in the propagator can shift the coefficient of the rz term (which is already O(p/Λ_b) near the boundary) by O(1), thereby changing g'/g and breaking the claimed matching to the gauge-invariant action. The paper's own summary flags the locality and information-loss issues (Section 7) but does not flag this substitution as an uncontrolled approximation. The boundary Ward-identity check in Appendix F verifies only the conformal form of the 3-point function, not the specific ratio that Eq. (44) produces.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper constructs a bulk AdS_4 action for a triplet of spin-1 fields from the Exact Renormalisation Group flow of the free USp(2N) vector model. The authors start from the flipped ERG equation for the SU(2) singlet currents, interpret the evolution operator as a D+1-dimensional path integral, map the radial variable to the Poincaré coordinate z, and fix the field redefinition through a Bessel-function regulator so that the quadratic term becomes the standard AdS vector action. They then compute the cubic term by performing a momentum integral with a simplified low-energy propagator, and show that, on-shell and in radial gauge with transverse boundary fields, the resulting action is classically equivalent to the cubic part of Yang-Mills plus an F^3 term, with couplings g = 1/(√N 128πγ^3) and g' = i g/6 (Eq. (60)). Appendix F checks that the boundary 3-point function computed from the final action has the form required by the conformal Ward identities of [47].","tokens_in":27708,"tokens_out":22481,"duration_ms":192539,"significance":"If correct, this is a notable first-principles derivation: the ERG flow of a boundary vector model determines not only the quadratic kinetic term but also the cubic Yang-Mills and F^3 couplings, up to the boundary two-point normalization γ, with the resulting bulk action reproducing boundary correlators by construction. The paper is careful in several respects: the quadratic term is mapped to the standard AdS vector action using a regulator fixed through Eq. (19); the final on-shell equivalence is checked by explicit computation in Section 6; and the boundary 3-point function is benchmarked against the independent conformal Ward-identity result [47]. The only free parameter γ is an output of the boundary two-point function, and the couplings are not inserted by hand. These strengths make the claimed result worth publishing if the main computational gap described below is closed.","major_comments":[{"comment":"The central claim of the paper depends on the k-integral in Eq. (44), which fixes the relative coefficient of the z^2 q_μ r_ν p_ρ term and the (1+rz) δ_{μν} p_ρ terms. This integral is not evaluated with the actual low-energy propagator Δ_l required by the regulator constraints discussed in the footnote to Section 5.2; instead, Δ_l is replaced by a 'simple' propagator, with the statement that the error in correlators is O(p/Λ_b). As written, this error estimate is asserted rather than derived, and it is an estimate for boundary correlators, not for the off-shell coefficients of the bulk action. Because the rz term in Eq. (44) is itself O(p/Λ_b) near the boundary, an O(p/Λ_b) error in the propagator can change the relative weight of the terms that ultimately determine g'/g through Eqs. (57)-(60). The authors should either compute the integral with the actual regulator or prove the O(p/Λ_b) bound at the level of the tensor coefficients in Eq. (44), including their z-dependence.","section":"Section 5.2, Eqs. (43)-(44)"},{"comment":"The check in Appendix F verifies that the boundary 3-point function computed from the final action has the tensor structure required by the conformal Ward identities of [47]. It does not verify the specific value of the ratio g'/g = i/6, nor the normalization in terms of γ, because the Ward identities determine the form of the correlator up to constants. Thus Appendix F is consistent with the central claim but does not provide independent evidence that the simplified-propagator computation in Eq. (44) produces the correct relative coefficients. A direct evaluation of the SU(2) current 3-point function in the free USp(2N) model, or an argument that the ratio g'/g is independent of the propagator choice, would close this gap.","section":"Appendix F"},{"comment":"The sentence claiming that the quantities outside the ∂_z parentheses are z-independent is not correct as written: the factors p^{-ν}/K_ν(pz), q^{-ν}/K_ν(qz), and r^{-ν}/K_ν(rz) in Eq. (43) depend on z, as do the bulk fields a_μ(p,z). If the text is taken literally, the integration by parts that leads to evaluation at the endpoints would drop the Bessel functions that later cancel against Eq. (44). The calculation in Appendix E appears to keep this z-dependence, so the main derivation should be rewritten to describe the actual manipulation consistently.","section":"Section 5.2, after Eq. (43)"}],"minor_comments":[{"comment":"There are minor typos: 'cublic' in the Section 5.2 heading, and 'verfied' and 'th ERG' in Section 7.","section":"Section 5.2 heading and Section 7"},{"comment":"For ν = 1/2, the numerical factor 2^{-1+ν} Γ(ν) equals √(π/2), not √π/2 as written; please correct this and verify the downstream constants in Eq. (55) and Appendix F.","section":"Eq. (42)"},{"comment":"The on-shell solution in Eq. (141) appears to drop the factor γ 2^{1-ν}/Γ(ν) present in Eq. (55); please clarify whether this is a redefinition of the source A_i^μ and ensure that Eq. (143) contains all constants when compared with [47].","section":"Appendix F, Eq. (141)"},{"comment":"The abstract and Introduction describe the cubic term as being 'those of' a gauge-fixed gauge-invariant action; Section 6 establishes this equivalence only on-shell, in radial gauge, and with transverse boundary fields, so the statement should be qualified as an on-shell/classical equivalence.","section":"Abstract and Introduction"},{"comment":"The appearance of the imaginary unit in g' = i g / 6 is surprising and deserves a brief comment on its origin, for instance whether it is a convention artifact of the Euclidean continuation used in Section 2.","section":"Eq. (60)"}],"recommendation":"major_revision","confidential_remarks":"For the editor: the paper is within the journal's scope and does not appear to be circular in its use of γ or the Ward-identity check. The main risk is that the simplified-propagator substitution in Section 5.2 is uncontrolled; if the authors can supply a derivation or a numerical/analytic check of Eq. (44) against the full regulator, the paper should be publishable. I do not see a need for additional novelty checks beyond normal editorial process."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should read this one. The paper derives the non-abelian cubic self-interaction of the bulk spin-1 field from the free USp(2N) vector model's ERG flow, and shows it matches, on-shell at cubic order, the gauge-fixed Yang-Mills plus F^3 vertex with g' = ig/6. That is a real first for this program: previous work only covered abelian spin-1, scalars, or scalar-graviton couplings. The boundary 3-point function computed from the final action is checked against the independent conformal Ward identity result [47], and the appendices are detailed enough that a patient reader can re-derive the main steps.\n\nThe weak point is exactly where the stress-test puts it. Section 5.2 replaces the complicated low-energy ERG propagator with a 'simple' propagator, asserting that the error in correlators is O(p/Lambda_b). The tensor structure of Eq. (44) — including the z-dependent terms that later combine into the YM and F^3 vertices — comes from that simplified propagator. The assertion is about correlators, not about the coefficients of the action, and an O(p/Lambda_b) error in the propagator can shift relative weights of the terms in Eq. (44) by O(1) after the on-shell replacement. The ratio g'/g depends on those relative weights, so the claim of matching the gauge-invariant action is only as solid as this unproven step. The paper's own summary flags the locality and information-loss issues but not this substitution, and the abstract omits the 'on-shell at cubic order' qualification, which is a bit stronger than what the body actually shows.\n\nThat said, the structural argument is sound and the paper is honest about its limitations elsewhere. The citation pattern is normal for a program paper: heavy self-citation, but the key external benchmark [47] is independent and used as a check, not to fix parameters. I do not think there is a circularity problem.\n\nWho should read it: anyone working on holographic RG, ERG-to-AdS maps, or vector-model/higher-spin duality. It deserves a serious referee. My recommendation: send it to peer review with the expectation that the authors address the simplified-propagator issue — either by proving the error estimate for the action coefficients or by finding a regulator that makes the integral tractable exactly — and that they tone down the abstract to match the on-shell, cubic-order statement.","headline":"A genuine proof-of-principle result for the ERG-to-AdS program, but the cubic vertex rests on an asserted regulator approximation that needs to be justified before the g' = ig/6 ratio is trusted.","tokens_in":28369,"tokens_out":2228,"would_cite":true,"duration_ms":21730,"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 boundary CFT's exact RG flow generates a Yang–Mills cubic interaction in the bulk.","keywords":["exact renormalisation group","holographic renormalisation group","Yang-Mills theory","AdS/CFT correspondence","vector model","spin-1 gauge field","conformal Ward identities","bulk reconstruction"],"falsifier":"Evaluate the cubic momentum integral (44) using the full regulator $I(p,z)$ from Eq. (41) rather than the simplified propagator, and compare the coefficients of $q^\\mu r^\\nu p^\\rho$, $\\delta^{\\mu\\nu}p^\\rho$, $q^\\mu\\delta^{\\nu\\rho}$, and $r^\\nu\\delta^{\\mu\\rho}$; if the ratios among these coefficients differ from those in Eq. (44), the on-shell equivalence to the gauge-invariant action and the relation $g'=ig/6$ fail. A second check is to compute the boundary three-point function from the bulk action at next order in $1/N$ and compare directly with the free-field current correlator of the $USp(2N)$ model.","tokens_in":27150,"feed_emoji":"⚛️","tokens_out":9022,"duration_ms":74355,"temperature":0.7,"pith_summary":"This paper derives a non-abelian spin-1 bulk action in AdS$_4$ from the exact renormalisation group (ERG) flow of a three-dimensional CFT, namely the $USp(2N)$-singlet sector of $2N$ free complex scalars. Its central claim is that the quadratic and cubic terms of the bulk action obtained as the ERG evolution operator are, once reduced on-shell and gauge fixed, those of a local gauge-invariant action: a Yang–Mills term quadratic in the field strength plus a term cubic in field strengths, with couplings $g = 1/(\\sqrt{N}\\,128\\pi\\gamma^3)$ and $g' = ig/6$. Because the bulk action is by construction the evolution operator for the boundary ERG equation, it is guaranteed to reproduce the correct boundary current correlators by the usual AdS/CFT prescription. If the claim holds, bulk gauge symmetry is not put in by hand but emerges from the renormalisation group of a boundary vector model.","feed_headline":"ERG flow of a vector CFT yields Yang–Mills in AdS4","feed_subtitle":"The cubic vertex matches gauge-fixed F² plus F³ with fixed coupling ratio g′=ig/6.","key_machinery":"The central object is the flipped Polchinski ERG equation for the $USp(2N)$-singlet spin-1 current, whose solution is written as a $D+1$-dimensional path-integral evolution operator; that operator defines the bulk action. A field redefinition $\\sigma^\\mu = z\\,f(z,p)\\,a^\\mu$ with $f$ fixed by a Bessel-function differential equation brings the kinetic term into the standard AdS form and maps the RG time $t$ to the Poincaré coordinate $z=1/\\Lambda$. The load-bearing calculation is the momentum integral in Eq. (44), performed with a simplified low-energy propagator, which converts the Bessel factors in the cubic term into $|p|$-like factors; those factors are exchanged for $z$-derivatives and $p^2$ factors using the classical field equation, and the resulting action is shown to equal the gauge-fixed form of the $F^2+F^3$ action.","core_discovery":"The paper shows that the cubic interaction produced by the RG flow of the spin-1 singlet current is classically equivalent, on-shell and in radial gauge with transverse boundary fields, to the cubic vertex of $S' = \\frac{1}{4}\\int d^{D+1}x\\,\\sqrt{G}\\,F^i_{MN}F^{i\\,MN} + g'\\epsilon_{ijk}\\int d^{D+1}x\\,\\sqrt{G}\\,F^{i\\,L}_{\\ \\ M}F^{j\\,M}_{\\ \\ N}F^{k\\,N}_{\\ \\ L}$, with Yang–Mills coupling $g=1/(\\sqrt{N}\\,128\\pi\\gamma^3)$ and $g'=ig/6$. The paper further verifies that the boundary three-point function computed from this action satisfies the conformal Ward identities in the form given by the momentum-space analysis it cites. Thus the ERG flow of the boundary theory is claimed to determine both the kinetic and cubic self-interaction of the bulk spin-1 gauge field, including the relative strength of the $F^3$ term.","pith_inferences":["A sharper test of the mechanism would be to compute the $O(1/N)$ terms of the same flipped ERG equation and check whether they generate the quartic Yang–Mills vertex with coefficients consistent with the same gauge-invariant completion.","Because the simplified propagator is justified by an asserted $O(p/\\Lambda_b)$ error, one could numerically evaluate the full Bessel-regulated momentum integral in Eq. (44) and see whether the relative coefficients that produce $g'=ig/6$ survive away from the strict $\\Lambda_b\\to\\infty$ limit.","If the same derivation is applied to the spin-2 current of the model, one would expect the gravitational cubic coupling to be fixed by the same boundary RG data, connecting this construction to the graviton-scalar cubic interaction already derived in the same line of work."],"forward_implications":["The ERG flow of the $USp(2N)$ vector model fixes not just the Yang–Mills coupling but also the strength of the $F^3$ cubic vertex, with the fixed ratio $g'=ig/6$.","Because the bulk action is the ERG evolution operator, boundary correlators computed from it automatically match the boundary theory; the paper confirms this explicitly for the spin-1 current three-point function and its conformal Ward identities.","The construction extends the holographic-RG programme from Abelian gauge fields and scalar or graviton-scalar couplings to a non-abelian, self-interacting gauge field.","The equivalence is established on-shell in the large-$N$ semiclassical limit, so the gauge-invariant action and the ERG-derived action give the same boundary correlators in that limit."],"supporting_citations":[{"why":"Supplies the prescription that turns a boundary ERG equation into a D+1-dimensional bulk action via the evolution operator.","marker":"[1]"},{"why":"Established the cubic scalar interaction and the regulator choice in the same holographic-ERG construction.","marker":"[31]"},{"why":"Provides the field redefinition and Bessel-function solution that put the spin-1 kinetic term into standard AdS form.","marker":"[32]"},{"why":"Shows how locality and general coordinate invariance are handled in this ERG-to-holographic-RG map, including the flipped ERG interpretation.","marker":"[33]"},{"why":"Identifies the vector model with a bulk higher-spin theory, the setting in which the spin-1 gauge field lives.","marker":"[34]"},{"why":"Gives the momentum-space conformal Ward identities used to verify the boundary three-point function.","marker":"[47]"},{"why":"Provides the exact renormalisation group equation on which the flipped ERG equation is based.","marker":"[48]"},{"why":"Supports the choice of the USp(2N) singlet sector and the su(2) adjoint structure of the odd-spin currents.","marker":"[61]"}],"fun_headline_variants":["Boundary ERG flow fixes Yang-Mills cubic vertex","RG flow of vector CFT yields YM interaction","AdS4 Yang-Mills from boundary ERG flow","Cubic F^3 coupling fixed by CFT flow"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The computation of the cubic vertex in Section 5.2 replaces the actual Bessel-function low-energy propagator with a simpler propagator, and the claim that this changes correlators only at $O(p/\\Lambda_b)$ is asserted rather than derived; if that simplification distorts the relative coefficients in the momentum integral (44), the match to gauge-fixed Yang–Mills plus $F^3$, and hence $g'=ig/6$, would not follow.","fun_headline_variants_meta":{"raw":{"variants":["Boundary ERG flow fixes Yang-Mills cubic vertex","RG flow of vector CFT yields YM interaction","AdS4 Yang-Mills from boundary ERG flow","Cubic F^3 coupling fixed by CFT flow"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000256,"raw_usage":{"total_tokens":1544,"prompt_tokens":886,"completion_tokens":658,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":502,"completion_tokens_details":{"reasoning_tokens":592}},"tokens_in":502,"tokens_out":658,"duration_ms":5521,"temperature":1.0,"reasoning_tokens":592,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T11:50:38.330782+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Evaluate the cubic momentum integral (44) using the full regulator $I(p,z)$ from Eq. (41) rather than the simplified propagator, and compare the coefficients of $q^\\mu r^\\nu p^\\rho$, $\\delta^{\\mu\\nu}p^\\rho$, $q^\\mu\\delta^{\\nu\\rho}$, and $r^\\nu\\delta^{\\mu\\rho}$; if the ratios among these coefficients differ from those in Eq. (44), the on-shell equivalence to the gauge-invariant action and the relation $g'=ig/6$ fail. A second check is to compute the boundary three-point function from the bulk action at next order in $1/N$ and compare directly with the free-field current correlator of the $USp(2N)$ model.","supporting_citations":[{"cited_title":"Holographic RG and Exact RG in O(N) Model","cited_arxiv_id":"2005.10412","evidence_quote":"Established the cubic scalar interaction and the regulator choice in the same holographic-ERG construction."},{"cited_title":"A Holographic form for Wilson's RG","cited_arxiv_id":"1706.03371","evidence_quote":"Supplies the prescription that turns a boundary ERG equation into a D+1-dimensional bulk action via the evolution operator."},{"cited_title":"Bulk Gauge Fields and Holographic RG from Exact RG","cited_arxiv_id":"2201.06240","evidence_quote":"Provides the field redefinition and Bessel-function solution that put the spin-1 kinetic term into standard AdS form."},{"cited_title":"Renormalization and Effective Lagrangians,","cited_arxiv_id":null,"evidence_quote":"Provides the exact renormalisation group equation on which the flipped ERG equation is based."}],"review_version":1}