{"id":"c4ef3229-3da5-4b56-96aa-bdf01263efe5","arxiv_id":"2501.00197","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The Delta Conjecture symmetric function is realized as the bigraded Frobenius character of the Borel-Moore homology of a new family of affine Springer fibers.","lead":"This paper constructs new geometric spaces, called affine Delta-Springer fibers, whose Borel-Moore homology encodes the symmetric functions behind the Delta Conjecture and the Rational Shuffle Theorem. It gives a geometric home for known combinatorial identities and points toward new geometric tools in this area.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 5.13 rests on the unproved rational smoothness assertion in Lemma 5.12; if the locus ev(C''∩Gr_γ) is not contained in the smooth locus of O_y, the W^P action does not transfer and Theorem 1.6(b) does not follow.","rationale":"The paper gives a coherent and detailed geometric program: X_{n,k} is paved in Section 6, the Springer action on X_{n,k} is standard, and the final identification with the Delta Conjecture symmetric function is a combination of the Rational Shuffle Theorem with the authors' earlier skewing formula. The most delicate point is not the paving but the transfer of the W^P action from the Borho–MacPherson sheaf to the Borel–Moore homology of Y_{n,k}. This transfer is exactly where the paper invokes rational smoothness of O_y over the relevant locus. Lemma 5.12 is the sole justification, and it is asserted rather than proved in this text; the implication from the Jordan-type bound of Lemma 5.10 to the pointwise rational smoothness of O_y on ev(C''∩Gr_γ) is not written out. Since every subsequent conclusion of Theorem 5.13—and hence Theorem 1.6(b)—depends on that implication, this is the most load-bearing assumption in the paper. I do not see an internal contradiction or an obvious counterexample, so the reader's CONDITIONAL verdict is appropriate; my concern reinforces the same weakest assumption rather than changing the verdict.","tokens_in":35294,"tokens_out":20498,"duration_ms":214871,"concrete_test":"For the smallest nontrivial cases (n,k)=(3,2),(4,2),(4,3), verify the pointwise hypotheses of [11, Lemma 3.4] on the locus ev(C''∩Gr_γ): enumerate the Jordan types occurring on the fibers, and for each type compute the local intersection cohomology stalk of O_y at a point of ev(Y_{n,k}). If any stalk is not one-dimensional in the expected degree, Lemma 5.12's smooth locus is too small and Theorem 5.13 fails; if all stalks match, re-derive the resulting graded S_n character of H^BM_*(Y_{n,k}) and compare it directly with rev_qωΔ'_{e_{k-1}}e_n in every bidegree.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing step is the construction of the S_n action on H^BM_*(Y_{n,k}). Theorem 5.13 proves the geometric skewing formula by restricting the Borho–MacPherson decomposition (27) to the locus Ξ and replacing IC(Q_{O_y}) by Q_{O_y}|_Ξ. That replacement is justified only by Lemma 5.12, whose entire proof is the sentence: \"This follows from Lemma 5.10 and [11, Lemma 3.4].\" Lemma 5.10 bounds the number of Jordan blocks of γ|_{Λ_0/ϵΛ_0} by k, but this is a statement about finite-dimensional quotients of the affine flag variety. Rational smoothness of O_y over the full image ev(C''∩Gr_γ) is a pointwise statement about the local intersection cohomology of O_y; the paper does not show that the hypotheses of [11, Lemma 3.4] hold at every such point of the affine Borho–MacPherson variety. If the smooth locus is strictly smaller than ev(Y_{n,k}), equation (28) is false, the W^P action does not descend to the restriction, and Theorem 1.6(b) is unsupported. This is not a disagreement with the result; it is the one step where the argument depends on an asserted, rather than demonstrated, geometric input.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces two families of subvarieties of (partial) affine flag varieties, X_{n,k,N} and Y_{n,k,N}, associated to a nil-elliptic operator γ_{n,k,N}, and proves that for N ≥ k their Borel-Moore homology, equipped with a Springer-type symmetric group action and a bigrading, has graded Frobenius character equal to rev_q ω(E_{K,k}·1) and rev_q ω(Δ'_{e_{k-1}}e_n), respectively. Here E_{K,k}·1 is the rational-shuffle symmetric function and Δ'_{e_{k-1}}e_n is the Delta Conjecture symmetric function, with K = k(n−k+1). The argument combines a combinatorial bijection between rational parking functions and γ-restricted affine permutations (Sections 2–3), a geometric construction of the varieties (Section 4), a sheaf-theoretic skewing formula proved via Borho–MacPherson theory (Section 5), and an explicit affine paving of X_{n,k} (Section 6). Theorem 1.6(b) is obtained by combining the geometric skewing formula (Theorem 5.13) with the authors' prior combinatorial skewing formula [10, Theorem 1.1].","tokens_in":35523,"tokens_out":26016,"duration_ms":247675,"significance":"If the main results are correct, the paper provides a genuine geometric interpretation of the Delta Conjecture and of the Rational Shuffle Theorem in the integer-slope case (km,k), generalizing Hikita's affine Springer fiber realization of ∇e_n. The affine paving of X_{n,k} is explicit and detailed, and the relation between the two geometric objects is mediated by a Schur skewing operator, giving a geometric avatar of the algebraic skewing formula. The combinatorial sections are thorough and the proof of Theorem 1.3 is essentially self-contained modulo standard facts. The main caveat is that the geometric transfer in Theorem 5.13 rests on a rational-smoothness assertion that is currently not demonstrated, so the full strength of the geometric realization depends on closing that gap.","major_comments":[{"comment":"Lemma 5.12 is the unique justification for replacing IC(Q_{O_y}) by Q_{O_y}|_Ξ in the proof of Theorem 5.13, yet its proof is the single sentence 'This follows from Lemma 5.10 and [11, Lemma 3.4].' Lemma 5.10 bounds the number of Jordan blocks of γ|_{Λ0/ϵΛ0} by k for points of Y_{n,k}; this is a statement about finite-dimensional quotients of the affine flag variety. Rational smoothness of O_y at the points in the preimage of ev(Y_{n,k}) is a pointwise statement about the local geometry of the affine Borho–MacPherson variety, and the paper does not verify that the hypotheses of [11, Lemma 3.4] hold at each such point. The issue is load-bearing: if the smooth locus of O_y does not contain the relevant fibers, equation (28) is false, the W^P action does not transfer, and Theorem 1.6(a), hence Theorem 1.6(b), is unsupported. Remark 5.7 explicitly notes that O_y is not rationally smooth on all fibers over ev(Gr_γ), so a global argument cannot be intended; the restriction to ev(C''∩Gr_γ) is essential. Please provide a complete pointwise verification, or quote the precise statement from [11] and check all of its hypotheses in detail.","section":"Section 5.2, Lemma 5.12"}],"minor_comments":[{"comment":"In the displayed computation of Hilb_{q,t}H^{BM}_*(pr_η(X_{n,k})), the factors q and t are written in opposite orders in two consecutive lines: one line has q^{area(π)}t^{δ_{K,k}−dinv′(π)} and the next has t^{area(π)}q^{δ_{K,k}−dinv′(π)}. These are not equivalent, and the q/t convention should be fixed so that the final equality with rev_qω(E_{K,k}·1) is correct.","section":"Section 6.2, proof of Theorem 1.6"},{"comment":"The notation for the partial affine flag variety is inconsistent: the text uses both eFl(K−n,1n) and fFl(K−n,1n) for the same object; please unify the notation.","section":"Theorem 1.6(b) and Definition 4.11"},{"comment":"The equivalence 'JT(Λ0/ΛK−n) ≤ (n−k)^{k−1} iff b_i ≤ n−k for all i' is asserted without proof; a short justification, even a sentence explaining the correspondence between Jordan blocks of the induced operator and columns of big labels, would improve readability.","section":"Lemma 4.14"},{"comment":"In the final lines of the proof, the notation E_{k,K} is used where the paper elsewhere defines E_{K,k}; please correct the order of the subscripts.","section":"Section 6.2, proof of Theorem 1.6"},{"comment":"The title in the manuscript body appears with misplaced spaces ('INTERPRET A TION', 'DEL T A'); please ensure the final formatting is correct.","section":"Title and running header"}],"recommendation":"major_revision","confidential_remarks":"The main technical obstacle is the rational-smoothness assertion in Lemma 5.12; this is likely fixable but needs to be supplied in detail. The manuscript also depends on the authors' unpublished preprint [10] for the key algebraic skewing formula; if the journal requires results to be independently available, the status of [10] should be verified. The paper is within the journal's scope and, apart from the noted gap, the argument is coherent and well structured."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The headline: this is a genuine step forward. The paper defines a new affine Delta-Springer fiber Y_{n,k} and proves Theorem 1.6(b), identifying its Borel-Moore homology (with its S_n action and bigrading) with rev_q ω Δ'_{e_{k-1}} e_n. If correct, this is the first geometric realization of the Delta Conjecture symmetric function, and it also covers the integer-slope Rational Shuffle case. The construction of Y_{n,k} and the geometric skewing formula (Theorem 5.13) are new; the earlier finite-dimensional Delta-Springer fibers did not reach this.\n\nThe paper does a lot well. Theorem 1.3's affine paving of X_{n,k} by (K,k) parking functions is a substantial, carefully proved technical achievement. Section 6 is detailed: the equations cutting out X∩C_ω are handled with a weight-ordering argument that is convincing. The overall strategy—pave X, transfer the Springer action via Borho–MacPherson, then apply the algebraic skewing formula from the authors' previous paper [10]—is coherent and mostly well executed.\n\nThe soft spots are real but localized. The one I would push on in review is Lemma 5.12. The rational smoothness of O_y over ev(C''∩Gr_γ) is asserted in a single sentence citing Lemma 5.10 and [11, Lemma 3.4]. This is load-bearing: it justifies replacing IC(Q_{O_y}) by Q_{O_y}|_Ξ in equation (28), and without it the W^P action does not transfer and Theorem 5.13 (hence 1.6(b)) does not follow. The stress-test note is right that the proof as written is not checkable without [11] in hand. A referee should ask for an expanded argument or a precise reference to the exact condition in [11, Lemma 3.4] verified at each point. I suspect it is fixable—Lemma 5.10 bounds the Jordan type length, which is exactly the kind of condition that controls rational smoothness—but the current one-line proof is too terse for such a central step.\n\nSecond, the introduction says the authors give a new combinatorial proof of the Rise Delta Conjecture. The body does not do that: Theorem 1.6(b) is a geometric realization, and a combinatorial proof would require a paving of Y_{n,k} or a direct stacked-parking-function expansion, neither of which appears. That is an overstatement in the intro, not a flaw in the main theorem.\n\nThe dependence on [10] for the key skewing identity is heavy, but that paper is independently argued and the identity is concrete; this is legitimate citation, not a circularity.\n\nWho this is for: algebraic combinatoricists and geometric representation theorists working on shuffle theorems. It deserves a serious referee. My recommendation is to send to peer review with a request that Lemma 5.12 be expanded or precisely referenced. The main geometric claim is likely correct; the gap is a missing verification, not a contradiction.","headline":"Genuine geometric realization of the Delta Conjecture via a new affine Springer-like fiber; the main gap is a terse rational-smoothness lemma that needs a fuller proof.","tokens_in":36140,"tokens_out":3809,"would_cite":true,"duration_ms":36735,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["05E05","05E10","14M15","14F43"],"pacs":[],"model":"deepseek-v4-flash","headline":"The authors build a variety $Y_{n,k}$ whose Borel-Moore homology, with an $S_n$ action and bigrading, has graded Frobenius character $\\mathrm{rev}_q\\,\\omega\\Delta'_{e_{k-1}}e_n$, matching the Delta Conjecture.","keywords":["Delta Conjecture","affine Springer fibers","Rational Shuffle Theorem","Borel-Moore homology","parking functions","Schur skewing operator","affine flag varieties","graded Frobenius character"],"falsifier":"Compute the graded character of $Y_{n,k}$ for the smallest case outside the known Shuffle Theorem diagonal, say $(n,k)=(5,3)$, by explicitly affine-paving the variety, and compare the Schur expansion with $\\mathrm{rev}_q\\,\\omega\\Delta'_{e_2}e_5$; any mismatch in a single $(q,t)$-degree would refute Theorem 1.6(b). A sharper check is the rational-smoothness premise itself: in the $(5,5)$ example of Remark 1.11, verify that $O_y$ is rationally smooth over the Schubert cell $C_{[1,7,8,9,15]}$ even though the fiber over the torus fixed point changes Springer type from $(3,1,1)$ to $(2,2,1)$.","tokens_in":35030,"feed_emoji":"🔺","tokens_out":13109,"duration_ms":115233,"temperature":0.7,"pith_summary":"This paper proves that the $\\Delta$ Conjecture, a central combinatorial formula for the symmetric function $\\mathrm{rev}_q\\,\\omega \\Delta'_{e_{k-1}}e_n$, is a statement about the homology of an explicitly defined variety. The authors introduce the affine $\\Delta$-Springer fiber $Y_{n,k}$, cut out from an affine Springer fiber by a union of Schubert cells and a Jordan-type bound, and show that its Borel-Moore homology carries an $S_n$ action and a bigrading whose graded Frobenius character is exactly that symmetric function. A companion variety $X_{n,k}$ gives the same geometric treatment of the $(K,k)$ Rational Shuffle Theorem, and a geometric skewing formula transfers one character to the other. If correct, the construction makes the $\\Delta$ Conjecture a concrete homology computation and a source of representation-theoretic structure behind the formula.","feed_headline":"Delta Conjecture realized by homology of a new Springer fiber","feed_subtitle":"Borel-Moore homology of a new variety carries an S_n action whose character is exactly the Delta Conjecture polynomial.","key_machinery":"The load-bearing objects are two subvarieties of (partial) affine flag varieties built from the nil-elliptic operator $\\gamma=\\gamma_{n,k,N}$ with characteristic polynomial $z^K-\\epsilon^{N+k}$: $X_{n,k,N}=\\operatorname{Sp}_\\gamma\\cap C$, where $C$ is the union of positive normalized Schubert cells, and $Y_{n,k,N}=BM_{\\gamma,n,k}\\cap C'$, where $BM_{\\gamma,n,k}$ adds the Jordan-type condition $\\operatorname{JT}(\\gamma|_{\\Lambda_0/\\Lambda_{K-n}})\\le (n-k)^{k-1}$ and $C'$ is its image in the partial affine flag variety $\\widetilde{\\operatorname{Fl}}^{(K-n,1^n)}$. The argument runs on two engines: the Springer action, supplied by the partial resolution of the nilpotent cone and the Decomposition Theorem, and an explicit affine paving of $X_{n,k}$ whose cell equations are triangular with respect to a $\\mathbb{C}^*\\times\\mathbb{C}^*$ weight order, so the cell dimensions compute the $\\operatorname{dinv}$ statistic of parking functions. The bridge between the two characters is the Schur skewing operator $s^\\perp_\\mu$, the adjoint of multiplication by a Schur function, whose geometric avatar is Theorem 5.13.","core_discovery":"The central discovery is Theorem 1.6(b): for $\\lambda' = (n-k)^{k-1}$, after the stabilization $N\\ge k$ the Borel-Moore homology of $Y_{n,k}=Y_{n,k,N}$ is a bigraded $S_n$-module with $\\operatorname{Frob}_{q,t} H_*^{BM}(Y_{n,k}) = \\mathrm{rev}_q\\,\\omega(\\Delta'_{e_{k-1}}e_n)$, where $q$ records homological degree and $t$ the connected component of the partial affine flag variety. The proof realizes this as the geometric counterpart of the skewing formula $\\Delta'_{e_{k-1}}e_n = s^\\perp_{(k-1)^{n-k}}(E_{K,k}\\cdot 1)$: Theorem 1.6(a) gives $q^{\\binom{k-1}{2}(n-k)}\\operatorname{Frob}_{q,t}H_*^{BM}(Y_{n,k}) = s^\\perp_{\\lambda'}\\operatorname{Frob}_{q,t}H_*^{BM}(X_{n,k})$, while $X_{n,k}$ admits an affine paving whose cells are labeled by $(K,k)$ parking functions and whose character is $\\mathrm{rev}_q\\,\\omega(E_{K,k}\\cdot 1)$. In the case $n=k$ the two varieties coincide and the statement specializes to the geometric Shuffle Theorem for $\\nabla e_n$.","pith_inferences":["This construction plausibly extends to other non-coprime rational slopes: replacing the Jordan-type bound by the analogue for a partition $\\mu$ should give geometric models for $E_{km,kn}\\cdot 1$, with $Y_{n,k}$ the case $(k(n-k+1),k)$.","The triangular cell equations and the $\\mathbb{C}^*\\times\\mathbb{C}^*$ weight order suggest an explicit monomial basis of $H_*^{BM}(Y_{n,k})$ indexed by stacked parking functions, which would give a combinatorial proof of Schur positivity of $\\Delta'_{e_{k-1}}e_n$ directly from geometry.","The $t$-grading by connected components together with the torus weights $\\theta_0,\\theta_\\infty$ may make $Y_{n,k}$ a geometric home for Delta-Conjecture analogues of double coinvariant modules, extending the $\\nabla e_n$ story; one test would be to compare the equivariant localization formula for $Y_{n,k}$ with the stacked parking function statistics."],"forward_implications":["The Delta Conjecture symmetric function $\\mathrm{rev}_q\\,\\omega\\Delta'_{e_{k-1}}e_n$ is the graded Frobenius character of an $S_n$-equivariant homology group, so every coefficient in its Schur expansion is a multiplicity of an irreducible $S_n$-representation on $H_*^{BM}(Y_{n,k})$.","The affine paving of $X_{n,k}$ gives $H_*^{BM}(X_{n,k})$ a cell basis indexed by $(K,k)$ parking functions, with the $\\operatorname{dinv}$ statistic appearing as cell dimension; the same cell structure carries over to the projection in the partial flag variety.","Theorem 1.6(a) supplies a geometric version of the skewing formula: the $S_n$-character of $Y_{n,k}$ is obtained from the $S_K$-character of $X_{n,k}$ by taking a $V_{(n-k)^{k-1}}$-isotypic component, up to a $q$-shift.","For $n=k$, $Y_{n,n}=X_{n,n}$ and the statement specializes to a geometric realization of the Shuffle Theorem for $\\nabla e_n$.","Since $Y_{n,k,N}$ is independent of $N$ for $N\\ge k$, the construction provides stable geometric models; varying $0\\le N