{"id":"394e2794-970c-4784-8306-bfd2bfe70bd7","arxiv_id":"2609.04035","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":8.0,"correctness_risk":"low","formal_verification":"none","parameter_count":1,"one_line_summary":"Cusp scaling operators transform as conformal primaries, and the Cusp Operator Expansion expresses any defect shape as a sum over cusp eigenstates.","lead":"This paper proves that cusped defect lines in any conformal field theory create operators that transform like ordinary primary operators under conformal symmetry, and that any smooth defect can be expanded in a basis of such cusped states. The results provide a new framework for analyzing defects with corners and open the door to bootstrap constraints on the data of these operators.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek/deepseek-v4-flash","headline":"Proof of primary transformation (3.8) assumes that under a conformal map, the regulator radius transforms via (3.5) to leading order; this relation depends on the regulator shape, which is not preserved by the map, threatening universality.","rationale":"The reader identified the C1 invariance assumption (3.2) as weakest; the present stress-test deepens that into a specific technical gap in the proof of (3.8): the transformation rule for the regulator radius (3.5) is not guaranteed for a generic C1 smoothing. The paper provides a consistency check in a free scalar theory, but only with the circular smoothing, leaving the universality claim unverified in that example. A concrete test with a different regulator shape would either confirm the claim or expose a limitation. Given the paper's strengths and the existence of some independent evidence (the free-scalar check), conditional acceptance is appropriate, requiring either clarification of the argument or an additional check. The verdict adjustment from ACCEPT to CONDITIONAL reflects this residual concern.","tokens_in":45863,"tokens_out":4513,"duration_ms":41583,"concrete_test":"Compute the two-point function of cusp operators (almond configuration) in the free scalar theory of Section 4.1 using a non-circular smoothing, e.g., replacing the circular cap by a small parabolic or square-shaped cap that joins the straight branches with C1 continuity. Renormalize by dividing out the leading divergence as in (3.4) and check whether the resulting correlator scales as 1/|x12|^{2Gamma} with the same Gamma given in (4.3). If the scaling exponent changes, or if the finite part acquires a shape-dependent factor that cannot be absorbed into operator normalization consistently, then the primary transformation law is not universal and the proof in Section 3.1.1 has a gap.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim that cusp operators transform like primaries is proven in Section 3.1.1 via a regularization that rounds each cusp with a circle of radius epsilon. Under a conformal map M, the rounded contour is assumed to map to a contour that is again a cusp rounded by circles of radii epsilon'_i = epsilon_i |∂x'_i/∂x_i|^{1/d} + O(epsilon^2). However, the image of a circle under a generic M is not a circle; the new contour is not obtained by the same rounding prescription. The authors claim universality (Section 3.1.1, paragraph 'Comments on Universality'), arguing that any C1 smoothing gives the same leading divergence and that scheme dependence only affects normalization. But this universality is not rigorously proven for a general CFT and relies on the unverified assumption that the relation (3.5) holds for the chosen definition of epsilon. Without this, the transformation law (3.8) does not follow from (3.6) because the renormalized correlator defined by (3.4) may depend on the shape of the regulator in a way that spoils the Jacobian factor in (3.5). The free scalar check in Section 4 uses only circular smoothing; it does not test the universality claim. This is a load-bearing gap because the entire covariance argument hinges on the relation (3.5) holding for the regularization scheme.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek/deepseek-v4-flash","summary":"The paper studies cusped line defects in conformal field theories.  It defines cusp operators via radial quantization of the Hilbert space around a cusp, introduces the Cusp Operator Expansion (COE), and proves that correlation functions of cusp operators on piecewise-circular contours transform like those of primary operators under general conformal transformations.  The results are checked in the free scalar pinning defect and used to derive spectral constraints via Tauberian analysis.  The central formal claim is that suitably defined cusp scaling operators are conformal primaries.","tokens_in":46294,"tokens_out":2403,"duration_ms":23301,"significance":"If the central claim holds, the paper provides a powerful new framework for studying cusped defects in any CFT.  The COE and the primary transformation law would imply that cusp correlation functions obey the same kinematic constraints as local primaries, opening the way to conformal bootstrap applications.  The explicit checks in the free scalar are non-trivial, and the spectral applications are interesting.  However, the proof of the primary transformation law has a load-bearing gap that needs to be addressed before the general claim can be accepted.","major_comments":[{"comment":"The proof that cusp operators transform like primaries uses a regularization that rounds each cusp with a circle of radius epsilon.  Under a conformal map M, the image of the rounded contour is assumed to be again a cusp rounded by circles of radii epsilon'_i = epsilon_i |\\partial x'_i/\\partial x_i|^{1/d} + O(epsilon^2).  However, the image of a circle under a generic conformal map is not a circle, so the regularization scheme is not shape-preserving.  The authors discuss universality (paragraph 'Comments on Universality'), arguing that any C1 smoothing yields the same leading divergence, but this universality is not proven for a general CFT.  The free scalar check only tests circular smoothing and does not verify the transformation of the cutoff under a generic map.  Since the entire derivation of Eq. (3.8) hinges on Eq. (3.5), this is a load-bearing gap.  The argument must be either strengthened by a rigorous proof of the universality assumption or replaced by a regularization that is manifestly covariant.","section":"3.1.1, Eq. (3.5) and surrounding proof"},{"comment":"The regularization for excited states (preparing exact eigenstates of the NS Hamiltonian) is conceptually different from the smoothing approach and may avoid the issue with image circles.  The authors should clarify whether this second construction indeed provides a covariant regularization independent of the smoothing scheme, and whether it can be used to give a self-contained proof of the primary transformation law without relying on Eq. (3.5).  As it stands, the paper presents two arguments, but the second one is also sketched briefly and its independence from the first is not explicit.","section":"3.1.2, excited states"}],"minor_comments":[{"comment":"The phrase 'the operator ¯Oa(p) change as ¯Oa(p)→Mba(p)¯Ob(p)' appears to have a typo: 'change' should be 'changes'.","section":"2.1"},{"comment":"The condition (2.9) is stated without derivation; a brief explanation or reference would improve accessibility.","section":"2.2.1, Eq. (2.9)"},{"comment":"The operator ˆO3 is defined as ϵ^2(∂_x ϕ - (2/x_2)ϕ)+c_2, but the coefficient c_2 is given later in Eq. (4.35) as 2λ cos(φ/2).  This mixing term depends on the coordinate x_2, which is a non-local feature.  The discussion around this point (page 33) is helpful, but a more explicit statement that this coordinate-dependence reflects the almond geometry would be welcome.","section":"4.3, Eq. (4.33)"},{"comment":"There is a missing factor in the exponent of the first line: 'e^{-\\Gamma y}' should presumably be 'e^{-\\Gamma y}' times the leading term; check the prefactor.","section":"Appendix C, Eq. (C.4)"},{"comment":"The notation 'N npoints' is unclear; define explicitly.","section":"3.2.3, Eq. (3.23)"},{"comment":"The relation of the Casimir energy E_φ to the cusp anomalous dimension is invoked but not derived.  A brief justification or reference would strengthen the argument.","section":"6.1"}],"recommendation":"major_revision","confidential_remarks":"The paper is well-organized and the physical intuition is clearly presented.  The free scalar example provides a valuable check.  However, the proof of the central claim (primary transformation law) is incomplete because it relies on an unverified universality assumption about the regularization scheme.  This gap is load-bearing: without it, the main result of Section 3 is not established.  I recommend that the authors either supply a rigorous argument for the universality of the cutoff transformation, or provide an alternative, manifestly covariant regularization.  If this can be fixed, the paper will be a significant contribution.  The current recommended action is major_revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek/deepseek-v4-flash","letter":"Dear colleague,\n\nThis paper aims to put cusped defects in any CFT on a systematic footing. It makes two main claims: (i) cusp scaling operators transform like primaries under conformal transformations, and (ii) the Cusp Operator Expansion (COE) lets us expand any defect shape in a basis of cusped states. Both are conceptually new and potentially important. The free-scalar examples (Section 4) are explicit and nontrivial; they give real credence to the COE machinery.\n\nMy reservation is with the proof of the primary transformation law (Section 3.1.1). The argument uses a circular smoothing of each cusp and asserts that under a conformal map the rounding scheme is preserved up to shape-independent normalization (the universality claim). The stress-test note is right: a generic conformal map does not send circles to circles, so the transformed contour is not obtained by the same smoothing prescription. The authors address this in the 'Comments on Universality' paragraph, arguing that any C1 smoothing gives the same leading divergence and that scheme differences only renormalise the cusp operator normalization. That could well be true, but it is not proven for a general CFT. The free-scalar check only uses circular smoothing; it does not test universality. So the covariance proof has a soft spot — not a fatal flaw, but a gap that a careful referee will want filled.\n\nThe COE itself and the worked examples are solid. The later bootstrap-inspired constraints (Section 6) are speculative but interesting. Overall the paper is a serious piece of work that deserves a proper referee. My own verdict is that the central claim is not yet airtight, but the framework is valuable enough to merit engagement.\n\nReading group: yes.\nWould cite: true.\nWould accept peer review: true.\nSerious thinker: yes.","headline":"A framework paper that rightly identifies cusp operators as primary-like objects, but the central proof of covariance rests on a regularization universality that is asserted rather than rigorously checked.","tokens_in":46798,"tokens_out":1732,"would_cite":true,"duration_ms":18072,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek/deepseek-v4-flash","headline":"Cusp scaling operators—excited states living at the corner of a defect line—transform like conformal primary operators, the paper proves, so cusped-contour correlators obey exactly the kinematics of local CFT correlators.","keywords":["cusped defects","conformal field theory","cusp operator expansion","cusp anomalous dimension","conformal bootstrap","defect conformal field theory","radial quantization","Tauberian analysis"],"falsifier":"Renormalize the two-cusp almond correlator of the pinning defect in an interacting variant of the theory (for instance, adding a quartic bulk coupling and working in $d=4-\\varepsilon$) using two different smoothings of the cusp, such as a circular cap and a non-circular cap. The paper predicts that the finite part after removing the $\\epsilon^{2\\Gamma}$ divergence is independent of the cap shape up to a frame-independent constant, equals the primary form $|x_1-x_2|^{-2\\Gamma}$, and shows no dependence on the bending parameter $\\chi$ that distorts the arcs without changing the cusp angle; any residual cap-shape or $\\chi$ dependence in the renormalized correlator would falsify the scheme-independence argument and the primary law built on it.","tokens_in":47686,"feed_emoji":"📐","tokens_out":12324,"duration_ms":104492,"temperature":0.7,"pith_summary":"This paper establishes that cusped line defects in any conformal field theory behave at each of their corners like conformal primary operators, despite the corner breaking the special conformal symmetries that primaries normally require. Concretely, a renormalized contour made of circular arcs meeting at cusps transforms under any conformal map by one Jacobian-power factor per cusp, with the cusp scaling dimensions $\\Gamma_{k_i}(\\varphi_i)$ playing the role of conformal dimensions and $\\varphi_i$ the opening angles. A sympathetic reader should care because this reduces two- and three-cusp correlators to fixed primary-style kinematics and makes four-cusp crossing equations a serviceable constraint on the cusp data—dimensions, OPE coefficients, and spectral densities—that appear in corner entanglement, static-source potentials, and gauge-theory cusp anomalies. The paper proves the claim by constructing the Hilbert space a cusp defines in radial and North-South quantization, derives an expansion of any orthogonally pierced contour piece in cusp states, verifies all of it in the magnetic-line defect of a free scalar in four dimensions, and then applies the expansion to derive Tauberian asymptotics for cusp spectral densities.","feed_headline":"Cusped defects obey primary-operator rules in any CFT","feed_subtitle":"Proof that cusp operators follow primary kinematics opens cusped Wilson loops and defects to conformal bootstrap.","key_machinery":"The cusp operator expansion (COE) is the load-bearing device: a piece of contour enclosed by a sphere that it pierces orthogonally is expanded in eigenstates of the dilatation operator, and any such state is a superposition of scaling cusp operators with dimensions $\\Gamma_n(\\varphi)$. For contours whose branches are arcs of circles, the same construction runs with the North-South Hamiltonian, the conformal image of dilatations whose flow lines are the arcs themselves, and the expansion of a rounded cusp of curvature radius $r$ has its coefficients fixed by symmetry to $c_n(\\varphi)\\,r^{\\Gamma_n}$. The proof of the primary law then feeds on one assumption—that two smooth ($C^1$) contours related by a conformal map have equal expectation values once renormalized—together with the way each small cutoff $\\epsilon_i$ rescales into $\\epsilon_i\\left|\\partial x'_i/\\partial x_i\\right|^{1/d}$; tracking this cutoff rescaling converts the covariance of the smoothed contour into the primary rescaling at each cusp.","core_discovery":"The central discovery is a transformation law, equation (3.1): for a piecewise-circular contour with states $\\left|k_i\\right\\rangle$ prepared at cusps $x_i$ with opening angles $\\varphi_i$, any conformal transformation acts only through factors $\\left|\\partial x'_i/\\partial x_i\\right|^{\\Gamma_{k_i}(\\varphi_i)/d}$ at each cusp, exactly as for local primaries, provided one keeps track of the transformed shape of the connecting arcs. Ground-state cusps are handled by softening each cusp into a circle of radius $\\epsilon$, isolating the leading $\\epsilon^{\\Gamma_0(\\varphi_i)}$ divergence through the cusp operator expansion, and defining the renormalized correlator as the finite, scheme-independent remainder; excited states are defined as eigenstates of the North-South Hamiltonian that leaves the two arcs meeting at the cusp invariant, which lets the same argument go through. The consequence is that planar two- and three-cusp correlators take the fixed primary forms $|x_1-x_2|^{-2\\Gamma}$ and $C(\\varphi_1,\\varphi_2,\\varphi_3)\\,x_{12}^{-\\Gamma_{123}}x_{23}^{-\\Gamma_{231}}x_{31}^{-\\Gamma_{312}}$, with all dynamics contained in the angle-dependent dimensions and coefficients, and that in the trivial-defect limit the excited cusp states degenerate into explicitly bi-local operators that still transform as primaries, resolving the tension with ordinary CFT where most operators are not primary.","pith_inferences":["I read the scheme-independence result as implying that the cusp dimensions $\\Gamma_n(\\varphi)$ are universal data of a CFT with defect, on the same footing as conformal dimensions, and therefore natural targets for measurement from the shape dependence of smooth contours; the paper itself stops short of stating this as a general principle.","Because the COE only requires the sphere to be pierced orthogonally, the same state-operator logic should extend to junctions of several defects and to higher-codimension objects; the paper mentions these generalizations but does not develop them.","The Section 5 construction shows that in a defect-free CFT the same mathematics packages any primary into a family of bi-local operators that still transform as primaries; I see this as a potentially reusable trick for building conformally covariant bases inside one-dimensional correlation functions, though the paper only uses it to demystify the cusp result.","The four-cusp crossing setup has a subtlety the paper flags: varying the cross ratio also changes the two virtual cusp angles on which the OPE data depend, so a numerical bootstrap would need to scan in the angle variables rather than the cross ratio; designing that scan is a concrete next step this paper leaves open."],"forward_implications":["Two- and three-cusp correlators on planar circular-arc contours are fixed up to normalization: $|x_1-x_2|^{-2\\Gamma}$ for the almond and $C(\\varphi_1,\\varphi_2,\\varphi_3)\\,x_{12}^{-\\Gamma_{123}}x_{23}^{-\\Gamma_{231}}x_{31}^{-\\Gamma_{312}}$ for the circular triangle, so all shape information beyond the cusp angles drops out.","An $n$-cusp planar contour carries exactly the residual kinematics of $n$ points in the plane once the cusp angles are fixed, so a four-cusp correlator can be decomposed in two COE channels and equality of the two channels is a genuine crossing equation on cusp data, which the paper argues opens the door to numerical bootstrap studies.","In the concentric-circle configuration, matching the COE to the fusion-limit Casimir divergence yields, via Kohlbecker's theorem, a Cardy-like exponential growth $\\exp(2\\sqrt{E_\\varphi \\Delta})$ for the integrated spectral density of cusp states weighted by squared COE coefficients.","For the fusion of two cusps into local defect operators, the same Tauberian strategy produces an asymptotic constraint on the weight $p_\\alpha(\\Delta)$ of the one-dimensional defect-CFT spectrum, with the ground-state cusp dimension entering the exponent of the growth.","The free-scalar magnetic-line example confirms the predicted two- and three-point forms, exhibits an excited-state tower whose dimensions are shifted by the cusp anomalous dimension, and reproduces a smooth contour's expectation value from the COE, checking the whole structure in closed form."],"supporting_citations":[{"why":"Supplies the ladder-limit observation that cusped contours behave like primary correlators in N=4 SYM and the first version of the cusp-cusp expansion that this paper generalizes to arbitrary conformal defects.","marker":"[40]"},{"why":"Provides the previous Ward-identity proof of ground-state cusp covariance for piecewise-circular Wilson loops, the direct precedent that the new proof extends to generic CFTs and excited states.","marker":"[44]"},{"why":"Foundational defect-CFT reference anchoring the assumption that smooth defect expectation values are conformally invariant.","marker":"[12]"},{"why":"Supports the C1 covariance property, equation (3.2), that renormalized smooth contours related by conformal maps have equal expectation values.","marker":"[69]"},{"why":"Gives the pinning-defect cusp anomalous dimension and the general convexity and unitarity constraints on the ground-state cusp dimension used throughout the worked examples.","marker":"[52]"},{"why":"Supplies the defect-fusion effective theory whose Casimir divergence, in the concentric-circle limit, is the input for the Tauberian constraint on the COE spectral density.","marker":"[62]"},{"why":"Kohlbecker's Tauberian theorem, used to convert the fusion-limit Laplace transform into the exponential growth of the integrated cusp-state density.","marker":"[83]"},{"why":"Tauberian method for conformal blocks, adapted in Section 6.2 to derive the weighted spectral-density asymptotics for the fusion of two cusps into defect operators.","marker":"[84]"}],"fun_headline_variants":["Cusp scaling operators transform as conformal primaries","Cusped defects obey primary kinematics, enabling bootstrap","Cusp operator expansion reveals primary-like transformation law","Cusps are primaries: new pathway for defect CFT bootstrap","Piecewise-circular cusp correlators take fixed primary form"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The proof rests on the assumption that any two smooth (continuously differentiable) defect contours related by a conformal transformation have exactly equal expectation values once UV divergences are renormalized, equation (3.2); if that covariance fails for general conformal maps, the primary transformation law for cusps fails with it.","fun_headline_variants_meta":{"raw":{"variants":["Cusp scaling operators transform as conformal primaries","Cusped defects obey primary kinematics, enabling bootstrap","Cusp operator expansion reveals primary-like transformation law","Cusps are primaries: new pathway for defect CFT bootstrap","Piecewise-circular cusp correlators take fixed primary form"]},"model":"deepseek/deepseek-v4-flash","effort":"low","cost_usd":0.000243,"raw_usage":{"cost":0.000242928,"is_byok":false,"cost_details":{"upstream_inference_cost":0.000242928,"upstream_inference_prompt_cost":0.000105168,"upstream_inference_completions_cost":0.00013776},"total_tokens":1448,"prompt_tokens":956,"completion_tokens":492,"prompt_tokens_details":{"audio_tokens":0,"video_tokens":0,"cached_tokens":256,"cache_write_tokens":0},"completion_tokens_details":{"audio_tokens":0,"image_tokens":0,"reasoning_tokens":424}},"tokens_in":956,"tokens_out":492,"cost_source":"openrouter_usage","duration_ms":5351,"temperature":1.0,"reasoning_tokens":424,"transport_provider":"openrouter","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-09-08T19:21:38.549457+00:00","model_set":{"reader":"deepseek/deepseek-v4-flash"},"falsifier":"Renormalize the two-cusp almond correlator of the pinning defect in an interacting variant of the theory (for instance, adding a quartic bulk coupling and working in $d=4-\\varepsilon$) using two different smoothings of the cusp, such as a circular cap and a non-circular cap. The paper predicts that the finite part after removing the $\\epsilon^{2\\Gamma}$ divergence is independent of the cap shape up to a frame-independent constant, equals the primary form $|x_1-x_2|^{-2\\Gamma}$, and shows no dependence on the bending parameter $\\chi$ that distorts the arcs without changing the cusp angle; any residual cap-shape or $\\chi$ dependence in the renormalized correlator would falsify the scheme-independence argument and the primary law built on it.","supporting_citations":[{"cited_title":"Quantum Spectral Curve and Structure Constants in N=4 SYM: Cusps in the Ladder Limit","cited_arxiv_id":"1802.04237","evidence_quote":"Supplies the ladder-limit observation that cusped contours behave like primary correlators in N=4 SYM and the first version of the cusp-cusp expansion that this paper generalizes to arbitrary conformal defects."},{"cited_title":"On anomalous conformal Ward identities for Wilson loops on polygon-like contours with circular edges","cited_arxiv_id":"2001.03391","evidence_quote":"Provides the previous Ward-identity proof of ground-state cusp covariance for piecewise-circular Wilson loops, the direct precedent that the new proof extends to generic CFTs and excited states."},{"cited_title":"Conformal Field Theories Near a Boundary in General Dimensions","cited_arxiv_id":"cond-mat/9505127","evidence_quote":"Foundational defect-CFT reference anchoring the assumption that smooth defect expectation values are conformally invariant."},{"cited_title":"Energy Momentum Tensor in Conformal Field Theories Near a Boundary","cited_arxiv_id":"hep-th/9302068","evidence_quote":"Supports the C1 covariance property, equation (3.2), that renormalized smooth contours related by conformal maps have equal expectation values."},{"cited_title":"Impurities with a cusp: general theory and 3d Ising","cited_arxiv_id":"2406.10186","evidence_quote":"Gives the pinning-defect cusp anomalous dimension and the general convexity and unitarity constraints on the ground-state cusp dimension used throughout the worked examples."},{"cited_title":"Effective theory for fusion of conformal defects","cited_arxiv_id":"2406.04561","evidence_quote":"Supplies the defect-fusion effective theory whose Casimir divergence, in the concentric-circle limit, is the input for the Tauberian constraint on the COE spectral density."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Kohlbecker's Tauberian theorem, used to convert the fusion-limit Laplace transform into the exponential growth of the integrated cusp-state density."},{"cited_title":"A tauberian theorem for the conformal bootstrap","cited_arxiv_id":"1709.00008","evidence_quote":"Tauberian method for conformal blocks, adapted in Section 6.2 to derive the weighted spectral-density asymptotics for the fusion of two cusps into defect operators."}],"review_version":1}