{"id":"26301a6d-9449-40d7-9378-1cab9258aebc","arxiv_id":"1908.06499","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For the affine Lie algebra A(2)2l, global Weyl modules are filtered by Demazure slices, the character of a Demazure slice is a nonsymmetric Macdonald-Koornwinder polynomial divided by its norm, and global Weyl modules for the special current algebra are free over their endomorphism ring.","lead":"Masahiro Chihara proves for the affine Lie algebra A(2)2l that global Weyl modules are filtered by Demazure slices and that the graded character of each slice is a nonsymmetric Macdonald-Koornwinder polynomial divided by its norm. This fills the one type-I case left open by Cherednik and Kato and adds a freeness theorem for global Weyl modules of the special current algebra.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Transfer of [CK] filtration/intersection lemmas to twisted type A(2) is asserted without proof; Theorem A's filtration and Theorem B's induction rest on it.","rationale":"The reader's weakest assumption is exactly the assertion that the Cherednik–Kato filtration and intersection lemmas transfer to type A(2)_{2l} without proof. This is the most load-bearing point in the paper: Theorem A's filtration of Gr_λD is built in Proposition 2.36 from Corollary 2.12, and Theorem B's proof uses the transferred Propositions 2.40 and 2.41. The paper gives substantial independent support elsewhere — Theorem 2.14 is quoted from Ion, the Weyl module character is tied to Koornwinder polynomials, and Theorem D is proved using Quillen–Suslin — but the central Demazure-slice filtration depends on unpublished-by-the-author transfer statements. I do not claim the transfer is false; the twisted root system has half-integral δ-shifts and a non-reduced finite part, so the burden is to show that the CK arguments survive those differences. A targeted rank-two computation of the Corollary 2.12 quotient is a feasible check. Because the concern does not by itself demonstrate a contradiction, it supports the reader's CONDITIONAL verdict rather than moving it; if the proposed check uncovers a counterexample, the verdict would need to move to REJECT.","tokens_in":26086,"tokens_out":7051,"duration_ms":72963,"concrete_test":"For l = 2, take λ = ε1 + ε2 and μ = ε1, so that λ ≻ μ by Definition 1.4(2) with λ+ − μ+ = ε2 ∈ ˚Q′_+. Set w = π_μ and v = π_λ, and compute the quotient (D_w ∩ D_v)/(D_v ∩ ∑_{u>w} D_u) in the level-one module L(Λ0) using the PBW basis of U(b−) and the realization of Theorem 2.5. If this quotient is neither 0 nor a copy of D_w, then Corollary 2.12 fails in type A(2)_{2l} and Proposition 2.36 must be reworked; if the quotient is D_w, the transfer is supported in this nontrivial example, though a full re-derivation of [CK, Cor. 4.4] with the twisted root sets would still be needed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 2.2 (after Lemma 2.8) states 'The proofs in [CK] are also valid for type A(2)_{2l}' for Lemma 2.9 and Corollary 2.12, but no proof or detailed translation is supplied. Lemma 2.9 controls inclusions D_v ⊆ D_w and minimal coset representatives, while Corollary 2.12 gives the intersection formula (D_w ∩ D_v)/(D_v ∩ ∑_{u>w} D_u) = D_w or 0. Proposition 2.36 uses Corollary 2.12 directly to produce the Demazure-slice filtration of Gr_λD, which is the content of Theorem A. The same unproved transfer underlies Proposition 2.40 ([CK] Prop 4.13) and Proposition 2.41 ([CK] Cor 4.15), which are then used in the inductive proof of Theorem 2.42 (Theorem B). The twisted root system differs materially from the untwisted case: real roots include 1/2(˚Δ_l + (2Z+1)δ), the Macdonald order uses ˚Q′ = ˚Q + (Z/2)˚Δ_l+, and the finite part is of type C_l. These differences are precisely where a 'proofs are also valid' assertion needs checking. If Corollary 2.12 fails for some pair w,v in A(2)_{2l}, the filtration in Proposition 2.36 loses its support, and with it the identification W(λ)⊗CΛ0 ≅ Gr_λD and the extension calculation Theorem 2.42.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies Demazure slices for the twisted affine Lie algebra of type A(2)_{2l}. It claims: (Theorem A, Theorem 2.38) the global Weyl module W(λ)⊗C CΛ0 is isomorphic to Gr_λ^D and is filtered by level-one Demazure slices, with each D_μ for μ in the ˚W-orbit of λ appearing exactly once; (Theorem B, Theorem 2.42) Ext^n_B(D_λ⊗C C_{mδ+kΛ0}, D^∨_μ) vanishes unless n=m=k=0 and λ=μ; (Theorem C, Corollary 2.44) the graded character of a Demazure slice equals a nonsymmetric Macdonald–Koornwinder polynomial divided by its norm; and (Theorem D, Theorems 3.15 and 3.16) for the special current algebra Cg†, the endomorphism ring End_{Cg†′}(W(λ)†) is a polynomial ring and W(λ)† is free over it. The proofs combine the Cherednik–Kato filtration lemmas, Demazure–Joseph functors, BGG resolutions, and results of Chari–Ion–Kus, Fourier–Kus, and Feigin–Makedonskyi.","tokens_in":26403,"tokens_out":11508,"duration_ms":101968,"significance":"If the main results are correct, this is a substantial extension of the Cherednik–Kato theory from untwisted affine algebras to the twisted type A(2)_{2l}, the case where the relevant nonsymmetric Macdonald polynomials are Koornwinder polynomials. The paper also proves a freeness result for global Weyl modules of the special current algebra that goes beyond existing dimension formulas. The manuscript is careful in its use of external benchmarks: the local Weyl module character formula, the Ext vanishing results of Chari–Ion–Kneser/Kleshchev, the dimension formulas of Fourier–Kus and Feigin–Makedonskyi, and the Quillen–Suslin theorem. There are no fitted parameters, and the central statements are checkable against these prior results. The main risk is the unproved transfer of the [CK] filtration and intersection lemmas to the twisted setting, which supports both Theorem A and Theorem B.","major_comments":[{"comment":"The statement \"The proofs in [CK] are also valid for type A(2)_{2l}\" is an unproved transfer of Lemma 2.9 and Corollary 2.12 from the untwisted setting. These results control inclusions D_v ⊆ D_w and the intersection formula (D_w ∩ D_v)/(D_v ∩ ∑_{u>w} D_u) = D_w or 0; Corollary 2.12 is used directly in Proposition 2.36 to obtain the Demazure-slice filtration, and the same transfer underlies Propositions 2.40 and 2.41 and the induction in Theorem 2.42. The twisted root system differs in exactly the places that matter: real roots include (1/2)(˚Δ_l + (2Z+1)δ), the lattice ˚Q′ = ˚Q + (Z/2)˚Δ_{l,+} is used in Lemma 1.5, and the finite part is of type C_l. Since [CK] is cited as an arXiv preprint rather than a published source, the reader cannot verify the transfer elsewhere. The author should either prove these lemmas for A(2)_{2l} or provide a detailed translation with the modified root system.","section":"Section 2.2 (after Lemma 2.8)"},{"comment":"The proof of Lemma 2.37 begins with Ext^k_{Cg-modint}(L(Λ0), (C_{−(λ|λ)δ/2} ⊗ W(λ)_{loc})^∨) = Ext^k(L(Λ0), D^∨_λ). Theorem 2.17 gives W(λ)_{loc} ≅ D_λ ⊗ C_{(λ|λ)δ/2−Λ0}, so the left-hand side involves D^∨_λ ⊗ C_{Λ0} (up to dual conventions), not simply D^∨_λ. The displayed equality is therefore not immediate, and the missing shift is material because it feeds into the character identity gch L(Λ0) = ∑_{λ∈˚P+} q^{(λ|λ)/2} gch W(λ), which is used in the proof of Theorem 2.38. Please spell out the twist or correct the displayed equality.","section":"Lemma 2.37, first displayed equality"},{"comment":"In the base case μ = 0, the proof asserts that for λ not anti-dominant, choosing i with s_i λ > λ gives Ext^n_B(D_i(D_λ ⊗ C_{mδ+kΛ0}), D^∨_0) = {0}, 'using Proposition 2.41'. Proposition 2.41 provides an exact sequence 0 → D_c → D_i(D_c) → D_{s_i c} → 0 and D_i(D_{s_i c}) = 0; it does not by itself imply that the middle object has vanishing Ext against D^∨_0. The argument needs an additional step, such as an explicit long exact sequence comparison or a further vanishing statement for D_{s_i λ}, and as written this step is missing. Since this is the starting point of the induction proving Theorem B, it should be completed.","section":"Theorem 2.42, base case of induction"}],"minor_comments":[{"comment":"The symbol q^{(b|b)/2} uses an undefined 'b'; it should presumably be q^{(λ|λ)/2} throughout the abstract, Theorem C/Corollary 2.44, and the statement of Theorem 2.14.","section":"Abstract and Theorems 2.14, 2.44"},{"comment":"There are several typos: 'Thorem 2.42' in the introduction, 'anaologues' for 'analogues', and in the proof of Corollary 2.44 the expansion is written with E†_λ in every summand, whereas the basis elements should be E†_μ / ⟨¯E_μ, E†_μ⟩.","section":"Introduction and proof of Corollary 2.44"},{"comment":"'Argumentation ideal' should be 'augmentation ideal' in the paragraph before Proposition 3.12; in the proof of Theorem 3.16, 'It is suffice' is a typo and '(W(λ))†)' has an unmatched parenthesis.","section":"Section 3"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is the missing A(2)_{2l} case of the Cherednik–Kato Demazure-slice program, and that is the main reason to read it. Cherednik and Kato explicitly left this type out, so the results are new: global Weyl modules filtered by Demazure slices, Ext vanishing between slices and thin Demazure modules, the character formula in terms of nonsymmetric Macdonald–Koornwinder polynomials, and freeness of the special current algebra Weyl module over its endomorphism ring. The paper does the work a paper in this area should do, and the last section on the special current algebra is especially clean.\n\nThe proofs are mostly detailed and honest. I saw no circular reasoning and no fitted parameters. The benchmarks are external: local Weyl module characters, the Chari–Ion–Kneser/Kleshchev vanishing theorem, Fourier–Kus dimension formulas, and Quillen–Suslin. The citation pattern is not a problem; the heavy reliance on [CK] and [CIK] is legitimate because those papers really do the neighboring cases.\n\nThe real soft spot is exactly the one the stress test flags. In Section 2.2, after Lemma 2.8, the paper says the proofs of [CK] Lemma 2.9 and Corollary 2.12 are also valid for type A(2)_{2l}, with no argument. Those lemmas control inclusions between Demazure modules and the intersection formula used to build the Demazure-slice filtration in Proposition 2.36. The twisted root system is materially different: real roots include half-root contributions, and the Macdonald order is built on ˚Q′ = ˚Q + (Z/2)˚Δ_l. So this is not a cosmetic comment. Is it fatal? I do not think so. The paper's own Lemma 1.6 does real work in adapting the order to A(2), and the combinatorics in [CK] is largely about the affine Weyl group, which is understood in this type. But a referee should ask for a proof or a precise reference with a translation, not a one-sentence assertion.\n\nThere is also a minor index typo in the proof of Corollary 2.44: the basis and expansion use E†_λ where E†_µ is clearly meant. That is easy to fix and does not affect the argument.\n\nWho is this for: specialists in current algebras, affine Demazure modules, and Koornwinder polynomials. It deserves a serious referee and, in my view, conditional acceptance. The central claims look right, but the CK transfer needs to be made explicit before I would sign off on Theorems A and B.","headline":"This is the missing A(2)_{2l} case of the Cherednik–Kato Demazure-slice program, done seriously and mostly concretely, with one unproved transfer of [CK] lemmas that a referee should force the author to spell out.","tokens_in":26936,"tokens_out":2481,"would_cite":true,"duration_ms":28510,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["17B67","17B10","33D52","17B65"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that for type $A_{2l}^{(2)}$ the global Weyl module is filtered by Demazure slices, that these slices are mutually orthogonal under the Euler–Poincaré pairing, and that their graded characters are normalized nonsymmetric…","keywords":["Demazure slices","thick Demazure modules","global Weyl modules","hyperspecial current algebras","special current algebras","nonsymmetric Macdonald-Koornwinder polynomials","Euler-Poincaré pairing","twisted affine type A(2)_{2l}"],"falsifier":"For a small rank, say $l=2$, compute the containment $D_v\\subseteq D_w$ for minimal coset representatives in the affine Weyl group of type $A_{2l}^{(2)}$ and check Lemma 2.9(1)–(2); if a containment occurs in the wrong Bruhat direction, or if the intersection quotient in Corollary 2.12 is neither $D_w$ nor zero, then the filtration in Proposition 2.36 and Theorem A collapse. Alternatively, compute $\\mathrm{Ext}^1_{\\mathcal{B}}(D_\\lambda\\otimes\\mathbb{C}_{m\\delta+k\\Lambda_0},D_\\mu^\\vee)$ for a specific nontrivial pair $(\\lambda,\\mu)$; Theorem B predicts that every such group vanishes except on the diagonal.","tokens_in":2597,"feed_emoji":"🧮","tokens_out":3069,"duration_ms":121803,"temperature":0.7,"pith_summary":"This paper extends to the twisted affine Kac–Moody algebra of type $A_{2l}^{(2)}$ results previously known for untwisted and other type-I affine algebras. Its central claim is that the global Weyl module of the hyperspecial current algebra, tensored with the level-one module $\\mathbb{C}_{\\Lambda_0}$, carries a filtration whose successive quotients are exactly the Demazure slices $D_\\mu$, one for each weight $\\mu$ in the finite Weyl group orbit of $\\lambda$ (Theorem A). A companion extension calculation (Theorem B) shows that slices and thin Demazure modules are mutually orthogonal under the Euler–Poincaré pairing, and this yields Theorem C: the graded character of a Demazure slice is a nonsymmetric Macdonald–Koornwinder polynomial specialized at $t=\\infty$, divided by its norm. In the final section the paper proves that for the special current algebra the global Weyl module is free over its polynomial endomorphism ring (Theorem D). If correct, the paper closes the $A_{2l}^{(2)}$ gap in the thick-Demazure-module story and shows that Koornwinder polynomials control the characters of these modules.","feed_headline":"Demazure slice characters are normalized Koornwinder polynomials","feed_subtitle":"For twisted type A(2)_{2l}, Weyl modules split into Demazure slices, one per orbit, with Koornwinder character formulas.","key_machinery":"The Demazure slice $D_w^\\Lambda:=D_w^\\Lambda/\\sum_{w<v}D_v^\\Lambda$ is the associated graded piece of a thick Demazure module, and at level one the paper writes $D_\\lambda$ for the slice attached to the minimal coset representative $\\pi_\\lambda$. The argument is carried by the identification $W(\\lambda)\\otimes\\mathbb{C}_{\\Lambda_0}\\cong\\mathrm{Gr}_\\lambda^D$, where $\\mathrm{Gr}_\\lambda^D=D_\\lambda/\\sum_{\\lambda\\succ\\mu,\\mu\\notin\\check W\\lambda}D_\\mu$; the Demazure–Joseph functors $D_i$ and their adjunction (Proposition 2.35) turn the extension calculation into an induction on the Macdonald order; and the nonsymmetric Macdonald–Koornwinder polynomials $E_\\lambda$ and their $t=\\infty$ specializations $E^\\dagger_\\lambda$ supply the characters, with the Euler–Poincaré pairing $\\langle-, -\\rangle_{\\mathrm{Ext}}$ serving as the inner product with respect to which slices and thin modules are dual.","core_discovery":"The paper's main discovery is that the missing twisted case behaves exactly like the cases already treated in the untwisted setting, with the half-integer root lattice of $A_{2l}^{(2)}$ causing no breakdown. For each dominant integral weight $\\lambda\\in \\check P_+$, the global Weyl module $W(\\lambda)\\otimes_{\\mathbb{C}}\\mathbb{C}_{\\Lambda_0}$ is isomorphic to the graded piece $\\mathrm{Gr}_\\lambda^D$ of the level-one Demazure module $D_\\lambda$, and as a $b^-$-module it is filtered by Demazure slices $D_\\mu$ with each $\\mu\\in\\check W\\lambda$ appearing exactly once. For every $\\lambda,\\mu\\in\\check P$, the extension groups $\\mathrm{Ext}^n_{\\mathcal{B}}(D_\\lambda\\otimes\\mathbb{C}_{m\\delta+k\\Lambda_0},D_\\mu^\\vee)$ vanish except for $n=m=k=0$ and $\\lambda=\\mu$, so the graded characters of slices and thin Demazure modules are orthonormal under the Euler–Poincaré pairing. The resulting character formula, $\\operatorname{gch}D_\\lambda=q^{(\\lambda|\\lambda)/2}E^\\dagger_\\lambda(x_1^{-1},\\ldots,x_l^{-1},q^{-1})/\\langle \\bar E_\\lambda,E^\\dagger_\\lambda\\rangle_{\\mathrm{Ext}}$, identifies each slice character as a normalized nonsymmetric Macdonald–Koornwinder polynomial at $t=\\infty$. For the special current algebra, the endomorphism ring $\\mathrm{End}_{(Cg^\\dagger)'}(W(\\lambda)^\\dagger)$ is a polynomial ring and $W(\\lambda)^\\dagger$ is free over it.","pith_inferences":["A natural extension, not pursued in the paper, would be to run the same transfer of filtration lemmas for the remaining twisted affine types not covered in the untwisted setting; the uniform Demazure–Joseph machinery suggests the filtration and character formulas should persist, possibly with a different half-integrality normalization.","The character formula can be examined in the limits $q\\to 1$ or as $t\\to\\infty$, which should yield explicit dimension formulas for Demazure slices that can be checked against the known dimensions of local Weyl modules.","The freeness theorem can be read as saying the global Weyl module is a flat family of local Weyl modules over a polynomial base; if the same holds for other special current algebras, then the finite-dimensional representation theory of those algebras would be governed by a commutative deformation space.","The isomorphism $W(\\lambda)\\otimes\\mathbb{C}_{\\Lambda_0}\\cong\\mathrm{Gr}_\\lambda^D$ suggests that tensor products or fusion products of Demazure slices could be studied through Weyl modules, connecting to the circle of ideas around fusion products and limit constructions."],"forward_implications":["For each dominant $\\lambda$, the graded character of $W(\\lambda)\\otimes\\mathbb{C}_{\\Lambda_0}$ is the sum of the slice characters $\\operatorname{gch}D_\\mu$ over $\\mu\\in\\check W\\lambda$, because each $D_\\mu$ appears exactly once in the filtration.","The orthogonality $\\langle\\operatorname{gch}D_\\lambda,\\operatorname{gch}D_\\mu\\rangle_{\\mathrm{Ext}}=\\delta_{\\lambda,\\mu}$ means the slice characters and the thin Demazure characters form a dual basis in the space of formal characters.","Theorem C gives a closed formula: every slice character is a normalized nonsymmetric Macdonald–Koornwinder polynomial at $t=\\infty$, so slice characters can be computed recursively from orthogonal polynomial data.","For the special current algebra, freeness of $W(\\lambda)^\\dagger$ over its polynomial endomorphism ring implies that every specialization at a maximal ideal has the dimension of the corresponding local Weyl module, reproducing the dimension formulas of Theorem 3.4."],"supporting_citations":[{"why":"Supplies the thick-Demazure containment and intersection lemmas (Lemma 2.9 and Corollary 2.12) that the paper transfers to type $A_{2l}^{(2)}$ and uses to build the slice filtration.","marker":"[CK]"},{"why":"Gives Theorem 2.14, the graded character of thin Demazure modules as nonsymmetric Macdonald polynomials at $t=0$, which the slice character formula is compared against.","marker":"[Ion]"},{"why":"Defines global and local Weyl modules for the hyperspecial current algebra and provides the identifications and extension vanishing used in Theorems 2.17–2.19 and Section 3.","marker":"[CIK]"},{"why":"Supplies the BGG-type extension vanishing for current algebra modules used in Theorem 2.19, whence Theorem 2.21.","marker":"[CI]"},{"why":"Contributes the affine highest-weight-category extension-vanishing result needed alongside [CI] for Theorem 2.19.","marker":"[Kle]"},{"why":"Provides the Demazure–Joseph adjunction (Proposition 2.35) used in the induction proving Theorem 2.42.","marker":"[FKM]"},{"why":"Gives the graded-character comparison under the Demazure–Joseph functor used in Theorems 2.34 and 2.39.","marker":"[Kas]"},{"why":"Gives the dimension formulas for local Weyl modules of the special current algebra used in Corollary 3.14.","marker":"[FK]"},{"why":"Provides, together with [Qui], the theorem that projective modules over polynomial rings are free, used in the proof of Theorem 3.15.","marker":"[Sus]"},{"why":"Provides the freeness theorem for projective modules over polynomial rings used in the proof of Theorem 3.15.","marker":"[Qui]"}],"fun_headline_variants":["Twisted A(2) Demazure slices match Koornwinder characters","Demazure slice characters: normalized Koornwinder for A(2)","Twisted case: Demazure slices yield Koornwinder formulas","A(2) slices: characters are Koornwinder polynomials","Demazure slices of twisted type A: Koornwinder characters"],"cache_read_input_tokens":28928,"weakest_assumption_plain":"The paper relies on the assertion that the filtration lemmas on Demazure-module containment and intersections, proved in the untwisted affine setting, remain valid verbatim for type $A_{2l}^{(2)}$, and this transfer is stated without proof.","fun_headline_variants_meta":{"raw":{"variants":["Twisted A(2) Demazure slices match Koornwinder characters","Demazure slice characters: normalized Koornwinder for A(2)","Twisted case: Demazure slices yield Koornwinder formulas","A(2) slices: characters are Koornwinder polynomials","Demazure slices of twisted type A: Koornwinder characters"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000186,"raw_usage":{"total_tokens":1370,"prompt_tokens":1035,"completion_tokens":335,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":651,"completion_tokens_details":{"reasoning_tokens":241}},"tokens_in":651,"tokens_out":335,"duration_ms":3314,"temperature":1.0,"reasoning_tokens":241,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:43:49.148854+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For a small rank, say $l=2$, compute the containment $D_v\\subseteq D_w$ for minimal coset representatives in the affine Weyl group of type $A_{2l}^{(2)}$ and check Lemma 2.9(1)–(2); if a containment occurs in the wrong Bruhat direction, or if the intersection quotient in Corollary 2.12 is neither $D_w$ nor zero, then the filtration in Proposition 2.36 and Theorem A collapse. Alternatively, compute $\\mathrm{Ext}^1_{\\mathcal{B}}(D_\\lambda\\otimes\\mathbb{C}_{m\\delta+k\\Lambda_0},D_\\mu^\\vee)$ for a specific nontrivial pair $(\\lambda,\\mu)$; Theorem B predicts that every such group vanishes except on the diagonal.","supporting_citations":[],"review_version":1}