{"id":"b8caee52-b8f3-47d4-8c6a-67a4ab23bc50","arxiv_id":"2505.03705","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Three- and four-point Coulomb branch amplitudes of N=4 SYM are expressed as symplectic Grassmannian integrals over SpGr(3,6) and SpGr(4,8), up to a fitted kinematic factor at four points.","lead":"The authors write the three- and four-particle scattering amplitudes of a massive version of N=4 super Yang-Mills theory as integrals over spaces called symplectic Grassmannians, and show the kinematic spaces of these amplitudes match those geometric spaces. If correct, it extends the modern geometry-first approach to scattering amplitudes from massless to massive particles.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The integral formulas are only evaluated in the frame ⟨i1i2⟩=-[i1i2], where they carry an unregulated δ(0); the promised generalization to independent angle and square brackets is not carried out, so the claimed description of the real-mass Coulomb branch amplitudes is not established.","rationale":"The reader's weakest_assumption already lists both the unperformed δ(0) generalization and the unproven n=4 kinematic-space dimension. I agree with the CONDITIONAL verdict, but I would rank the δ(0) issue first because it affects every configuration in the standard real-mass frame, whereas the rank degeneracy is a lower-dimensional subvariety. The paper contains substantial concrete checks: the appendix localizations are explicit, the Macaulay2 computation is cited, and the six-dimensional three-point rewriting is verified against [55]. Those are real evidence, and I do not see a concrete sign error in the displayed algebra. The load-bearing unsupported step is the sentence in Sec. 4 saying the analysis 'can easily be generalised': all subsequent formulas are written in the frame where the two mass-conservation delta functions coincide, making the equality singular. A focused re-derivation of the localization without imposing ⟨i1i2⟩=-[i1i2] would settle whether the proposed f3 and f4 actually reproduce the Coulomb branch amplitudes for real masses or only for a complexified/six-dimensional kinematic slice.","tokens_in":32436,"tokens_out":8328,"duration_ms":88403,"concrete_test":"Redo the Appendix D.1 localization in the general complex-mass frame with independent ⟨i1i2⟩ and [i1i2], using the Sec. 3.2 parametrization (3.15)–(3.17): solve δ(CΩΛ^T) δ(CΩeΛ^T) δ(CΩC^T) without imposing ⟨i1i2⟩=-[i1i2], and check whether the f3 of Eq. (4.6) still evaluates to 1 on the resulting C* and whether the integral gives δ(Σ⟨i1i2⟩)δ(Σ[i1i2])δ^4(P) with no δ(0). If yes, the promised generalization exists; if not, the displayed amplitude formulas are only valid in the singular frame.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central formulas (1.1), (4.3), and (4.10) are derived in the frame where ⟨i1i2⟩ = -[i1i2], i.e. the standard real-mass Coulomb-branch frame with m̃_i = m'_i = m_i. In that frame Eq. (4.2) leaves δ(Σ⟨i1i2⟩) δ(Σ[i1i2⟩) = δ(Σm_i) δ(-Σm_i), a product of two identical delta functions; after imposing mass conservation this contains δ(0)^2 and is not a finite equality. The authors acknowledge this in Sec. 4 ('Strictly speaking, this would imply the appearance of a δ(0)...') and state that the analysis 'can easily be generalised', but the generalisation is never performed. All explicit localizations — C* in (4.4), the minor identities (4.8)–(4.9), the evaluations f3|_{C*}=1 and f4|_{C*}=-2s13/(s12s23)^2, and the six-dimensional reduction — use ⟨i1i2⟩=-[i1i2]. Consequently the displayed equalities in Sec. 4 are formal in the physical real-mass frame. A secondary but related gap is that for n=4 the joint row space of Λ and eΛ is only 'expected' to be four-dimensional (Sec. 3.2), so the claimed equivalence of the kinematic space to SpGr(4,8) is not proven for degenerate configurations.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes a symplectic Grassmannian formulation of three- and four-point tree-level superamplitudes on the Coulomb branch of N=4 SYM. After reviewing the massive spinor-helicity formalism and the special three-body kinematics, the authors rewrite the three-point amplitude as a single supercharge-conserving delta function and propose a general SpGr(n,2n) integral ansatz in Eq. (4.1). They evaluate this integral for n=3 and n=4 in Appendix D, obtaining the amplitudes of reference [47] up to a kinematic factor, with integrands f3 and f4 chosen specifically for this purpose. The last part expresses six-dimensional N=(1,1) SYM amplitudes in four-dimensional variables, in a form suggestive of SpGr(3,6) and SpGr(4,8) structure, with a detailed dimensional reduction of spinor and supersymmetry variables in the appendices.","tokens_in":32841,"tokens_out":9227,"duration_ms":83222,"significance":"If the construction is valid in full generality, it would be a significant step toward a Grassmannian geometry for massive amplitudes, connecting the Coulomb branch of N=4 SYM to the six-dimensional symplectic Grassmannian program. The paper contains detailed and apparently correct localization computations in Appendix D, and it is commendably explicit about the non-uniqueness of the integrands and about the delta-function(0) issue in the real-mass frame. However, as discussed below, the central equalities are established only in a restricted frame and the n=4 kinematic-space identification is not proved, so the paper currently establishes an existence result in a special frame rather than a fully general formulation.","major_comments":[{"comment":"The integral evaluations that support the central claims are performed in the frame where angle brackets and square brackets satisfy the relation ⟨i1i2⟩ = -[i1i2]. In that frame, which is the real-mass Coulomb branch frame with m̃_i = m'_i = m_i, the two mass-conservation delta functions in Eq. (4.2) are not independent: after imposing the mass-conservation condition Σ m_i = 0, the product δ(Σ⟨i1i2⟩) δ(Σ[i1i2]) contains a factor δ(0)^2, so the equality in Eq. (4.2) is formal. The paper acknowledges this ('Strictly speaking, this would imply the appearance of a δ(0)...') and states that the analysis can easily be generalized, but the generalization is not carried out; the localization of C to C* in Eq. (4.4), the minor identities in Eqs. (4.8)-(4.9), and the evaluations f3|_{C*}=1 and f4|_{C*}=-2s13/(s12s23)^2 all use the restricted frame. The central equalities (4.3) and (4.10) therefore do not yet establish an integral representation for the physical real-mass Coulomb branch amplitudes; a general-frame evaluation, or an argument that the δ(0) factor cancels in a well-defined limit, is needed.","section":"Section 4 (Eqs. (4.2), (4.3), (4.10); Appendix D)"},{"comment":"The claimed equivalence of the four-point kinematic space to SpGr(4,8) rests on the joint row space of Λ and eΛ being four-dimensional. The text states only that this is 'expected' in Section 3.2 and provides no proof, nor does it discuss degenerate kinematics where the intersection of the two row spans is non-zero. Since Eq. (4.10) localizes C to the 4×8 matrix (Λ;eΛ), a lowering of the rank would make C fail to define a point of SpGr(4,8) and would make the constraints C.Ω.C^T=0 and C.Ω.Λ^T=0 overdetermined. A dimension-counting proof for generic kinematics and a treatment of the degenerate cases is required for the statement that the kinematic space is exactly SpGr(4,8).","section":"Section 3.2 and Section 4.2"},{"comment":"The integrands f3 and f4 are not derived; they are chosen to satisfy f3|_{C*}=1 and f4|_{C*}=-2s13/(s12s23)^2, and the kinematic prefactor -s12s23/(2s13) is inserted into f4 to cancel the factor produced by the integral. The authors explicitly note the non-uniqueness of f3 and call f4 'a (non-unique) choice'. This means the paper establishes the existence of a symplectic Grassmannian integral representation that reproduces the known amplitudes, but not a canonical or predictive formulation: the integrand depends on external kinematics for n=4, multiple choices exist for n=3, and no f_n is known for n>4. This is a limitation rather than a logical flaw, but the conclusions should state it clearly and should not present the representation as a derivation.","section":"Section 4.1 and 4.2"}],"minor_comments":[{"comment":"The claim that Derksen's algorithm found a single generator of the invariant ring, and that this generator vanishes under the symplectic ideal, should be backed by the generator itself or by an ancillary file; as written it cannot be checked from the text.","section":"Section 3.2"},{"comment":"The phrase 'can be written merely as a supercharge-conserving delta function' should be qualified: the rewriting in Eq. (1.1) holds in a specially chosen little-group frame and after dropping an overall sign or normalization.","section":"Section 1.1"},{"comment":"The notation for C*, \\bar C*, u versus \\tilde u, and w versus \\tilde w is easy to confuse; a summary table of definitions would improve readability.","section":"Section 5"},{"comment":"The square roots of m_i/m*_i require a branch choice; please state the phase convention explicitly.","section":"Appendix B, Eq. (B.10)"},{"comment":"There are typos such as 'Pl¨ cuker' and 'renders the above amplitude two important features'; a careful proofread is recommended.","section":"Section 6"}],"recommendation":"major_revision","confidential_remarks":"The main obstacle is the unperformed generalization of the localization to the case where ⟨i1i2⟩ and [i1i2] are independent, together with the missing proof of the four-dimensionality of the joint row space for n=4. If the authors can supply these missing pieces, or alternatively restrict the claims appropriately, I would support publication. The computational core in Appendix D is solid and the paper is honest about its limitations; I do not see a reason to reject."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a genuine but modest step toward a symplectic Grassmannian description of massive amplitudes, and the title overstates what is proven for the physical Coulomb branch. The two things actually worth keeping are the geometric derivation of the three-point special kinematics from the isotropy of the joint row space of Λ and eΛ, and the explicit SpGr(3,6) and SpGr(4,8) integral formulas with careful localization in Appendix D. The Macaulay2/Derksen observation that no little-group invariant Plücker combination survives the symplectic ideal is a clean, reproducible detail.\n\nThe soft spots are real, and the stress-test note lands. The central equalities (1.1), (4.3), (4.10) are evaluated in the frame ⟨i1i2⟩ = -[i1i2]. There the two mass-conservation deltas coincide, so the right-hand sides contain a δ(0) and are not finite distributional equalities. The authors see this and write that the analysis can easily be generalised, but the general solution without that frame condition is never displayed. So the paper does not actually establish the claimed equivalence for real-mass Coulomb branch kinematics; the well-defined version lives in the complex-mass / six-dimensional setting. That is the load-bearing gap, not a quibble.\n\nSecond, f3 and f4 are reverse-engineered, and the paper is honest about it: f3 is a non-unique function chosen to equal 1 on the support, and f4 carries explicit Mandelstam dependence plus a prefactor inserted to cancel what the integral produces. That keeps the representation from being a derivation. For low-point work in the Grassmannian program this is an acceptable starting point, but a principle for f_n—beyond reproducing known amplitudes—is still missing.\n\nThird, the n=4 kinematic-space equivalence rests on an expected dimension of the joint row space of Λ and eΛ. For generic kinematics that is four, but degeneracies are not discussed.\n\nThe citation pattern is fine: the prior 6D symplectic Grassmannian work [51-53] and the Coulomb branch amplitudes [47] are properly credited, and the new content is clearly differentiable from them. Net assessment: the three-point geometry is a nice, checkable result; the four-point integral is a promising ansatz with careful localization; the δ(0) issue and fitted integrands mean the physical claims need qualification. A serious referee should see this, and the paper should come back with either a full complex-mass treatment or a regulated handling of the real-mass limit. I would engage with it, and I would cite the three-point rewriting.","headline":"Genuine but modest step toward a symplectic Grassmannian description of massive amplitudes; the three-point geometry is elegant, but the headline claim overstates what is proven because the displayed equalities carry an unregulated δ(0) in the physical real-mass frame.","tokens_in":33345,"tokens_out":5266,"would_cite":true,"duration_ms":52982,"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":"The three- and four-point Coulomb branch amplitudes of N=4 super Yang-Mills theory are integrals over the symplectic Grassmannians SpGr(3,6) and SpGr(4,8).","keywords":["symplectic Grassmannian","Coulomb branch","N=4 super Yang-Mills","scattering amplitudes","on-shell functions","massive spinor-helicity","six-dimensional N=(1,1) SYM","superamplitudes"],"falsifier":"Compute the dimension of the joint row space of $\\Lambda$ and $\\tilde\\Lambda$ for a four-particle configuration in which, say, one particle's angle spinors are proportional to another particle's square spinors. If that dimension is less than four, no $C_{4\\times8}$ can satisfy $C\\Omega C^T=C\\Omega\\Lambda^T=C\\Omega\\tilde\\Lambda^T=0$, and the claimed equivalence to $\\mathrm{SpGr}(4,8)$ fails at that kinematic point. Alternatively, evaluate the four-point integral in a frame with $\\langle i_1i_2\\rangle\\neq-[i_1i_2]$, avoiding the $\\delta(0)$, and compare the result with the known amplitude.","tokens_in":32223,"feed_emoji":"📐","tokens_out":10775,"duration_ms":95589,"temperature":0.7,"pith_summary":"The paper claims that the three- and four-point scattering super-amplitudes on the Coulomb branch of $N=4$ super Yang-Mills theory are naturally integrals over symplectic Grassmannians: the three-point amplitude is exactly a supercharge-conserving delta function over $\\mathrm{SpGr}(3,6)$, and the four-point amplitude is the same type of integral multiplied by a known function of minors. The massive kinematic data, encoded in the angle and square spinor matrices $\\tilde\\Lambda$ and $\\Lambda$, are themselves isotropic two-planes, and for three and four particles their joint row space forms the $n$-plane that defines the symplectic Grassmannian. This matters because it extends the Grassmannian programme for massless amplitudes to massive theories and gives a geometric explanation of the special three-body kinematics and of the mass-conservation conditions. If the description is right, higher-point Coulomb branch on-shell functions should fit the same ansatz once the missing integrand $f_n$ is known.","feed_headline":"Massive scattering amplitudes become symplectic-Grassmannian integrals","feed_subtitle":"For N=4 SYM on the Coulomb branch, three- and four-point amplitudes live on SpGr(3,6) and SpGr(4,8).","key_machinery":"The central object is the symplectic Grassmannian $\\mathrm{SpGr}(n,2n)$: the space of $n$-planes in a $2n$-dimensional complex vector space that are isotropic for a skew form $\\Omega$, i.e. matrices $C_{n\\times 2n}$ with $C\\Omega C^T=0$. The argument is carried by the three linear conditions $C\\Omega C^T=0$, $C\\Omega\\Lambda^T=0$, and $C\\Omega\\tilde\\Lambda^T=0$, which package mass conservation, momentum conservation, and the relation between angle and square spinors, respectively; the delta functions in the ansatz (4.1) localize $C$ onto the particular matrix $C_*$ built from the external data. The $u$-variables of three-particle special kinematics enter as the parametrization of the one-dimensional intersection of the row spaces of $\\Lambda$ and $\\tilde\\Lambda$, and a Groebner-basis invariant-ring computation shows that no little-group-invariant polynomial of Pluecker coordinates survives on $\\mathrm{SpGr}(3,6)$, forcing the integral to be written in a special little-group frame.","core_discovery":"The paper's central claim is a symplectic-Grassmannian integral representation for Coulomb branch amplitudes at three and four points. For three particles the amplitude collapses to a single supercharge-conserving delta function, $A_3=\\delta^6(C_*\\,\\Omega\\,\\eta^T)$, where $C_*$ is a $3\\times6$ matrix assembled from spinor-helicity variables, $\\Omega$ is the $6\\times6$ symplectic form, and $C_*\\Omega C_*^T=0$ identifies $C_*$ with a point of $\\mathrm{SpGr}(3,6)$. For four particles the same delta function appears with a prefactor that is the inverse of a product of two maximal minors of the $4\\times8$ matrix $C_*=(\\tilde\\Lambda;\\Lambda)$, so the integral (4.1) evaluates to the amplitude up to a known kinematic factor, or exactly after an explicitly given rescaling. The paper also demonstrates that the kinematic space of these amplitudes is equivalent to $\\mathrm{SpGr}(n,2n)$, that the massless MHV and anti-MHV three-point amplitudes are recovered as limits of the massive formula, and that the three- and four-point amplitudes of six-dimensional $N=(1,1)$ SYM, written in four-dimensional variables, display the same symplectic structure.","pith_inferences":["If the dimension count generalizes, the same ansatz (4.1) should describe higher-point Coulomb branch on-shell functions; the undetermined $f_n$ would then be fixed by BCFW amalgamation of three-point blocks rather than by direct integration, as in the massless Grassmannian construction.","Kinematic loci where the row spaces of $\\Lambda$ and $\\tilde\\Lambda$ have a larger-than-generic intersection form a boundary in amplitude space; on that boundary the $\\mathrm{SpGr}(n,2n)$ parametrization must either degenerate or acquire extra data, which could be tested by studying near-collinear massive kinematics.","A concrete next test is to evaluate the four-point symplectic integral in a frame without the condition $\\langle i_1i_2\\rangle=-[i_1i_2]$, i.e. with complex masses; the paper promises this generalization, and matching the known $1/(s_{12}s_{23})$ amplitude there would confirm the $\\mathrm{SpGr}(4,8)$ identification beyond the $\\delta(0)$-carrying frame."],"forward_implications":["At three points the Coulomb branch amplitude is exactly the supercharge-conserving delta function $\\delta^6(C_*\\Omega\\eta^T)$; no separate prefactor is needed, and the $\\mathrm{SL}(3)$ invariance of the amplitude becomes manifest.","At four points the same integral ansatz yields the amplitude up to the known factor $-2s_{13}/(s_{12}s_{23})$; an explicitly chosen $f_4$ removes this factor and gives the amplitude exactly.","In the massless limit the $\\mathrm{SpGr}(3,6)$ description degenerates into the ordinary Grassmannian descriptions of the MHV and anti-MHV three-point amplitudes, recovering $\\mathrm{Gr}(2,3)$ and $\\mathrm{Gr}(1,3)$ inside the symplectic Grassmannian.","The three- and four-point amplitudes of six-dimensional $N=(1,1)$ SYM, expressed in four-dimensional massive variables, take the same $\\delta^4(C_*\\Omega\\eta^T)\\delta^4(C_*\\Omega\\tilde\\eta^T)$ form, showing the symplectic structure is not an artifact of the four-dimensional complex-mass framing."],"supporting_citations":[{"why":"Supplies the massive little-group-covariant spinor-helicity formalism that underlies all the kinematic variables used in the paper.","marker":"[38]"},{"why":"Constructs the three-point Coulomb branch superamplitude and its special BPS kinematics, which the paper repackages into symplectic-geometry language.","marker":"[47]"},{"why":"Gives the on-shell functions and BCFW bridge on the Coulomb branch, including the box-diagram comparison that the four-point integral is checked against.","marker":"[50]"},{"why":"Provides the prior symplectic-Grassmannian integral construction in six dimensions that motivates the four-dimensional Coulomb branch ansatz.","marker":"[53]"},{"why":"Defines the six-dimensional massless spinor-helicity variables used for the dimensional reduction in Section 5.","marker":"[54]"},{"why":"Supplies the six-dimensional supertwistor three-point amplitude form that the paper matches in Appendix E.","marker":"[55]"},{"why":"The massless Grassmannian formulation of N=4 amplitudes that the symplectic Grassmannian ansatz is modeled on.","marker":"[32]"}],"fun_headline_variants":["Coulomb branch amplitudes become symplectic Grassmannian integrals","SpGr integrals for three- and four-point amplitudes","Symplectic Grassmannians capture massive amplitudes","Three- and four-point amplitudes on symplectic Grassmannians"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The $\\mathrm{SpGr}(n,2n)$ identification assumes the joint row space of $\\Lambda$ and $\\tilde\\Lambda$ is exactly $n$-dimensional for $n=3,4$ — proven for three points, only asserted as \"expected\" for four — and the explicit integral is evaluated in the frame $\\langle i_1i_2\\rangle=-[i_1i_2]$, which introduces a $\\delta(0)$.","fun_headline_variants_meta":{"raw":{"variants":["Coulomb branch amplitudes become symplectic Grassmannian integrals","SpGr integrals for three- and four-point amplitudes","Symplectic Grassmannians capture massive amplitudes","Three- and four-point amplitudes on symplectic Grassmannians"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000319,"raw_usage":{"total_tokens":1807,"prompt_tokens":961,"completion_tokens":846,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":577,"completion_tokens_details":{"reasoning_tokens":779}},"tokens_in":577,"tokens_out":846,"duration_ms":7773,"temperature":1.0,"reasoning_tokens":779,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T23:46:05.361052+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the dimension of the joint row space of $\\Lambda$ and $\\tilde\\Lambda$ for a four-particle configuration in which, say, one particle's angle spinors are proportional to another particle's square spinors. If that dimension is less than four, no $C_{4\\times8}$ can satisfy $C\\Omega C^T=C\\Omega\\Lambda^T=C\\Omega\\tilde\\Lambda^T=0$, and the claimed equivalence to $\\mathrm{SpGr}(4,8)$ fails at that kinematic point. Alternatively, evaluate the four-point integral in a frame with $\\langle i_1i_2\\rangle\\neq-[i_1i_2]$, avoiding the $\\delta(0)$, and compare the result with the known amplitude.","supporting_citations":[{"cited_title":"Symplectic Grassmannians, dual conformal symmetry and 4-point amplitudes in 6D","cited_arxiv_id":"2204.10014","evidence_quote":"Provides the prior symplectic-Grassmannian integral construction in six dimensions that motivates the four-dimensional Coulomb branch ansatz."}],"review_version":1}