{"id":"c76b112d-d714-4839-8ccc-eb84d2d66928","arxiv_id":"2608.00428","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"This expository preprint proposes a Langlands-style conjectural description of associated varieties and simple modules of simple affine vertex algebras at integer levels in simply-laced types.","lead":"This expository paper lays out two conjectures describing the geometry and representation theory of simple affine vertex algebras at integer levels in simply-laced types, using nilpotent orbits, Kazhdan-Lusztig cells, and affine Springer fibers. It is a useful entry point to an emerging Langlands-style dictionary for vertex algebras.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The load-bearing bridge is Theorem 4.2.2: both conjectures inherit its cell-orbit correspondence, and its proof is deferred to a case-by-case verification; an independent check in an exceptional type is needed.","rationale":"The paper is an honest expository account of a conjectural program: Conjecture 5.1.1 is attributed to [SYZ25], Conjecture 5.2.1 is explicitly deferred to future work, and the examples are checked against independent known results. The reader's weakest_assumption correctly identifies Theorem 4.2.2 as the load-bearing cell-orbit bridge, and I agree with that identification. My stress-test adds no new mathematical contradiction but sharpens the concern: Theorem 4.2.2 is the one input without which neither the associated variety formula nor the simple-module formula can stand, and its proof is not contained in this preprint. The z(l)\\times u term is also a genuine heuristic, but it affects only Conjecture 5.1.1 at non-distinguished levels, so it is less central than the cell-orbit bridge. The proposed E6/E7 check is a concrete, falsifiable way to test whether the case-by-case verification in [SYZ25] has hidden exceptions; if it passes, the conditional acceptance is well supported, and if it fails, both central conjectures collapse. Since the reader already set the verdict to CONDITIONAL and this concern does not change that verdict, the recommendation is UNCHANGED.","tokens_in":16163,"tokens_out":32100,"duration_ms":312427,"concrete_test":"Run an independent computation for E6 and E7: from the definition of cl_n, list O(m) for 1\\le m\\le h; compute the dominant affine weight \\xi_m, its stabilizer W_m, the longest element w_m, and the two-sided cell c(w_m). Then verify the equivalent form of Theorem 4.2.2 given in [SYZ25, Lemma 4.2.3] and [Yun16, Proposition 4.8]: the j-induction j^{W_{aff}}_{W_m}\\mathrm{sgn} of the sign representation of W_m must be the Springer representation of (\\check O(m), \\mathrm{triv}). If this fails for any m, Theorem 4.2.2 and both conjectures are falsified; if it passes in E6 and E7, the main support gap is substantially closed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is Conjectures 5.1.1 and 5.2.1. Both inherit Theorem 4.2.2: the associated variety conjecture has nilpotent part d\\check O(m), and the simple-module conjecture identifies the answer with the affine left cell module of w_m. Theorem 4.2.2 is quoted from [SYZ25, Theorem 4.2.1], and its proof is described only as case-by-case, with the dominant weights \\xi_m computed for all 1\\le m\\le h. If \\check O(m) is not special in some simply-laced type, or if Lusztig's bijection sends \\check O(m) to a two-sided cell different from c(w_m), then both conjectures fail even though the A5 and D4 examples match. The z(l)\\times u term in Conjecture 5.1.1 is a second heuristic and affects only non-distinguished levels, whereas Theorem 4.2.2 is indispensable for both conjectures, including all distinguished cases covered by Conjecture 5.2.1. No internal inconsistency in the paper was found; the gap is that the single most load-bearing input is not proved in this text.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This expository paper proposes an affine analog of the Barbasch--Vogan construction for simple affine vertex algebras. For a simply-laced simple Lie algebra g and an integer level k with m = k + h^vee >= 1, the authors define a cyclotomic level map cl_n and orbits O(m), state a cell-orbit bridge (Theorem 4.2.2) from their companion work [SYZ25], and then formulate two conjectures: Conjecture 5.1.1 describes the associated variety X_{L_k(g)} as a saturation of data attached to a Bala-Carter Levi of O(m), and Conjecture 5.2.1, in the distinguished case, identifies the Grothendieck group of the relevant category O with an affine left-cell module at q=1. The paper also reviews the finite-dimensional primitive-ideal picture, the Kazhdan-Lusztig cell formalism, and the 4D mirror-symmetry motivation, and it gives supporting examples in types A5 and D4. The main theorems are not proved here but are quoted from the authors' preprint [SYZ25], with the proof of Theorem 4.2.2 described as case-by-case.","tokens_in":16454,"tokens_out":11606,"duration_ms":101904,"significance":"If the two conjectures hold, they give a uniform, parameter-free description of both the associated variety and the simple-module category of L_k(g) for all integer levels in simply-laced types, directly generalizing the classical relationship between primitive ideals, nilpotent orbits, and Kazhdan-Lusztig cells. The conjectures also connect to quasi-lisse vertex algebras, affine Springer fibers, and 4D mirror symmetry, so they are likely to be influential if confirmed. The paper is honest and careful: it explicitly marks what is conjectural, checks the conjectures against independent published results (AM18a, AFK24, JS25, Per13, GK07), and provides concrete tables. There is no sign of fitted constants or circular derivations; the main caveat is the heavy reliance on a case-by-case theorem from a companion preprint.","major_comments":[{"comment":"The statement \"cl_n^{-1}([1,m]) = O(m)\" cannot be literally correct, since the left-hand side is a set of nilpotent orbits while O(m) is defined to be a single orbit. The following sentence, comparing with fibers of the duality map d, indicates that the intended statement is that O(m) is the unique maximal element of this preimage. Because O(m) is the key input to Conjectures 5.1.1 and 5.2.1, the theorem should be restated precisely and either proved or explicitly located with a page/equation reference in [SYZ25].","section":"Section 4.1, Theorem 4.1.2"},{"comment":"Both main conjectures inherit the identification of the orbit O(m) with the two-sided cell c(w_m) from Theorem 4.2.2, yet the proof is described only as a case-by-case verification in the preprint [SYZ25]. For this paper to be self-contained enough for the conjectures to be credible, the authors should either reproduce the proof or give a detailed summary of the verification, including the computed dominant weights xi_m and at least one independent check in an exceptional type. Alternatively, the conjectures should be explicitly stated as conditional on the validity of Theorem 4.2.2 in all simply-laced types.","section":"Section 4.2, Theorem 4.2.2"},{"comment":"The formula X_{L_k(g)} = AdG . (d_L O_L x z(l) x u) is the central associated-variety conjecture, but the notation is ambiguous: the three factors are subsets of l, z(l), and u, and the intended operation should be the sum of subsets inside p = l + u, not a Cartesian product. This is not merely cosmetic, because the later computation of the nilpotent part uses the identification of the product with a subset of g; please clarify the notation in the statement of the conjecture.","section":"Section 5.1, Conjecture 5.1.1"}],"minor_comments":[{"comment":"Rows m=4 and m=5 both list O(m) = (53). Since Theorem 4.2.2 attaches a two-sided cell c(w_m) to O(m), the table may appear to contradict the injectivity of that correspondence; please add a sentence explaining that cl_n is not injective and that w_4 and w_5 are distinct elements of the same two-sided cell.","section":"Example 5.1.4"},{"comment":"The label \"Conjeture\" should be \"Conjecture\". Also, the dotted arrows from c(w_m) are not explained in the caption; a one-sentence description of the diagram conventions would help the reader.","section":"Figure 5.3.1"},{"comment":"The sentence \"The details of this subsection and the next will appear in a future paper\" makes it unclear which aspects of Conjecture 5.2.1 are expected and which are already established; please mark the conjecture as explicitly conditional on Theorem 4.2.2 and on the promised future work.","section":"Section 5.2"},{"comment":"The authors note that the dominant weights xi_m for all 1 <= m <= h \"seem to be of independent interest\" but do not list them; including a table of these weights, even in an appendix, would strengthen the expository value of the paper.","section":"Section 4.2"}],"recommendation":"major_revision","confidential_remarks":"The paper is an honest exposition of a promising framework, and I see no circularity or fitted parameters. My main concern is that the credibility of both central conjectures rests on a case-by-case theorem from the authors' own preprint, whose proof is not reproduced here, and on a theorem statement in Section 4.1 that appears to be misstated. These issues are fixable within the manuscript's scope. I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear X, quick take on 2608.00428. It's an honest expository paper about a pair of conjectures, not a new proof. The piece is worth reading for the synthesis: the classical Barbasch-Vogan picture in §2 is very clearly assembled, and the affine diagram (5.3.1) is a useful way to see the program. The genuinely new item is Conjecture 5.2.1 (simple modules via affine left cells), with details deferred to a future paper. It fits naturally with the known example due to Perše.\n\nWhat the paper does well: it states the load-bearing objects cleanly — cyclotomic level map, orbits O(m), cell bridge — and is candid throughout about what is proved versus conjectured. Conjecture 5.1.1 on associated varieties is concrete, and the examples in A5 and D4 match the cited literature. I also appreciate that no free parameters or fitted data appear; circularity risk is low.\n\nThe soft spots are real but proportionate. The entire package rests on Theorem 4.2.2, quoted from [SYZ25], whose proof is case-by-case and not reproduced. Both conjectures inherit that bridge. If the cell-orbit correspondence fails in some simply-laced type, both conjectures fail. So I would like to see an independent check in an exceptional type, or at least a detailed account of the verification. A second heuristic is the z(l) x u term in 5.1.1, coming from [AFK24]; it only affects non-distinguished levels, while 5.2.1 deals with distinguished ones, so the two conjectures have different exposure. The presentation itself is careful and I found no internal inconsistency.\n\nWho this is for: specialists in affine vertex algebras, W-algebras, and Kazhdan-Lusztig theory who want a map of this program. It is not a self-contained proof source.\n\nRecommendation: I would send it to peer review. A serious referee is warranted; the comments should push for more evidence for Theorem 4.2.2 in exceptional types and a clearer statement of what exactly is proved in [SYZ25]. The paper deserves to be published as an expository/conjecture paper, ideally with those gaps flagged.","headline":"Honest expository map of a conjectural program; the new Conjecture 5.2.1 and the affine diagram are worth having, but the load-bearing cell-orbit theorem is quoted from prior work and still lacks an independent check in exceptional types.","tokens_in":17037,"tokens_out":1847,"would_cite":true,"duration_ms":16795,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["17B69","17B67","20C08"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper conjectures that for every integer level in simply-laced type, the associated variety and simple modules of $L_k(\\mathfrak g)$ are governed by a single nilpotent orbit and a single left cell.","keywords":["affine vertex algebras","associated varieties","Kazhdan-Lusztig cells","nilpotent orbits","Barbasch-Vogan duality","cyclotomic level map","quasi-lisse vertex algebras","simple modules"],"falsifier":"Compute $X_{L_k(\\mathfrak g)}$ for a simply-laced type not in the paper's tables, for example type $E_6$ at $m=3$ ($k=-9$), and compare with $\\operatorname{Ad}G\\cdot(d_L O_{\\check L}\\times\\mathfrak z(\\mathfrak l)\\times\\mathfrak u)$; any discrepancy falsifies Conjecture 5.1.1. Independently, list $\\mathrm{cl}_n$ on all nilpotent orbits of $E_8$ and check the equality $\\check O(m)=\\mathbf c(w_m)$ through Lusztig's cell bijection; one mismatch would invalidate Theorem 4.2.2 and with it both conjectures.","tokens_in":15894,"feed_emoji":"📐","tokens_out":9231,"duration_ms":73437,"temperature":0.7,"pith_summary":"This expository paper proposes a uniform description of simple affine vertex algebras $L_k(\\mathfrak g)$ for a simple simply-laced Lie algebra $\\mathfrak g$ and integer level $k$ above the critical level. Writing $k+\\check h=m\\ge 1$, the authors conjecture that the associated variety of $L_k(\\mathfrak g)$ equals $\\operatorname{Ad}G\\cdot(d_L O_{\\check L}\\times \\mathfrak z(\\mathfrak l)\\times \\mathfrak u)$, where $\\check O(m)$ is the unique nilpotent orbit attached to $m$ by their cyclotomic level map, and that when $\\check O(m)$ is distinguished the Grothendieck group of the block $O_{\\xi_m-\\hat\\rho}(L_k(\\mathfrak g))$ is isomorphic to the affine Kazhdan-Lusztig left cell module $H^{\\mathrm{aff}}_{c_L(w_m)}$ at $q=1$. This is meant as an affine analog of the classical Barbasch-Vogan picture for primitive ideals, with $L_k(\\mathfrak g)$ playing the role of the quotient $U(\\mathfrak g)/J_{\\lambda,\\max}$. If correct, the two conjectures give one framework governing both the Poisson geometry and the simple module category for all integer levels in simply-laced types, and they pin down exactly when $L_k(\\mathfrak g)$ is quasi-lisse.","feed_headline":"One orbit and one left cell predict L_k(g) at every integer level","feed_subtitle":"If right, associated varieties and simple modules follow one rule in all simply-laced types.","key_machinery":"The carrying object is the cyclotomic level map $\\mathrm{cl}_n:\\mathcal N\\to\\mathbb Z_{\\ge 1}$, sending a nilpotent element $e$ to $\\min\\{m\\in\\mathbb Z_{\\ge1}\\mid(\\operatorname{ad}e)^{2m}=0\\}$. It selects, for each $m$, a unique nilpotent orbit $O(m)$ whose closure is $\\mathrm{cl}_n^{-1}([1,m])$. The companion object is the affine Weyl group element $w_m$, defined as the longest element of the stabilizer of the dominant translate $\\xi_m$ of $m\\Lambda_0+\\rho$ under the affine action. Theorem 4.2.2 is the bridge: under Lusztig's bijection between nilpotent orbits of the dual group and two-sided cells of the affine Weyl group, the orbit $\\check O(m)$ corresponds exactly to the two-sided cell $\\mathbf c(w_m)$. The conjectures then attach to $L_k(\\mathfrak g)$ the orbit data $d_L O_{\\check L}$, $\\mathfrak z(\\mathfrak l)$, $\\mathfrak u$, and the left cell $c_L(w_m)$.","core_discovery":"The central claim is that the simple affine vertex algebra $L_k(\\mathfrak g)$ at an integer level $k$ with $m=k+\\check h\\ge 1$ should be understood as the affine counterpart of the maximal primitive ideal quotient $U(\\mathfrak g)/J_{\\lambda,\\max}$. Conjecture 5.1.1 asserts an exact formula for the associated variety: $X_{L_k(\\mathfrak g)}=\\operatorname{Ad}G\\cdot(d_L O_{\\check L}\\times \\mathfrak z(\\mathfrak l)\\times \\mathfrak u)$, where $\\check O(m)=\\operatorname{Sat}_{\\check G}^{\\check L} O_{\\check L}$ is the unique orbit with cyclotomic level $m$, $d_L$ is Barbasch-Vogan duality for the Levi, $\\mathfrak z(\\mathfrak l)$ is the center of the Bala-Carter Levi, and $\\mathfrak u$ is the nilpotent radical of a matching parabolic. Conjecture 5.2.1 asserts that when $\\check O(m)$ is distinguished, the block of category $\\mathcal O$ for $L_k(\\mathfrak g)$ has Grothendieck group $H^{\\mathrm{aff}}_{c_L(w_m)}|_{q=1}$, with the irreducible objects bijecting with the left cell $c_L(w_m)$ of the affine Weyl group element $w_m$. The paper also states that a future work will add the matching affine Springer fiber side of the picture.","pith_inferences":["The cyclotomic level map is defined uniformly by a nilpotency index, so one could test Conjecture 5.1.1 by direct computation in types $E_6,E_7,E_8$, where the paper's tables do not go; a mismatch would show the z(l)-term or the orbit assignment is not the right invariant.","Because the left cell $c_L(w_m)$ is explicitly described in simply-laced affine Weyl groups, Conjecture 5.2.1 predicts exact character formulas for the new simple modules, which could be checked by computing Zhu algebra modules in a type not yet examined.","If the $\\mathfrak z(\\mathfrak l)\\times\\mathfrak u$ term is correct, the semisimple part of the associated variety depends only on the orbit $O(m)$, implying that $X_{L_k(\\mathfrak g)}$ is constant across levels sharing the same $O(m)$; this constancy is an implicit testable consequence.","The framework suggests a broader Langlands-style dictionary between nilpotent orbits in the dual group and representations of affine vertex algebras at rational levels, with the integer-level conjectures forming the first uniform case."],"forward_implications":["If Conjecture 5.1.1 holds, $L_k(\\mathfrak g)$ is quasi-lisse exactly when $\\check O(m)$ is distinguished, because the semisimple factor $\\mathfrak z(\\mathfrak l)$ vanishes precisely in that case.","The nilpotent part of the predicted associated variety is Lusztig-Spaltenstein induction $\\mathrm{Ind}_L^G d_L O_{\\check L}=d\\check O(m)$, so the entire geometry is controlled by the single orbit $O(m)$ and its Levi.","Conjecture 5.2.1 gives an explicit finite list of simple modules in the block: one for each element of the left cell $c_L(w_m)$, with the Grothendieck group identified with the affine dual left cell module at $q=1$.","The conjectures complete the affine analog of the classical diagram (2.5.1), adding an affine Springer fiber whose top cohomology should carry the right cell module (5.3.1).","All checked cases in types $A_5$ and $D_4$ agree with the conjectures, including previously known results for the vacuum and quasi-lisse levels."],"supporting_citations":[{"why":"Supplies the cyclotomic level map, the orbit O(m), and Theorem 4.2.2 identifying O(m) with the two-sided cell of w_m; both main conjectures are stated there.","marker":"[SYZ25]"},{"why":"Constructs the bijection between nilpotent orbits of the dual group and two-sided cells of the affine Weyl group on which Theorem 4.2.2 relies.","marker":"[Lus89]"},{"why":"Defines the Barbasch-Vogan duality d used in Conjecture 5.1.1 and in the equality d Sat = Ind.","marker":"[BV85]"},{"why":"Provides the Kazhdan-Lusztig map from nilpotent orbits to conjugacy classes and the fixed-point variety arguments behind the cyclotomic level map.","marker":"[KL88]"},{"why":"Establishes that minimal reduction type and the Kazhdan-Lusztig map coincide, identifying the cyclotomic level as the maximal cyclotomic factor of a Weyl group element.","marker":"[Yun21]"},{"why":"Supplies Lusztig-Spaltenstein induction of orbits, used to express the nilpotent part of the predicted associated variety as Ind_L^G d_L O_L.","marker":"[LS79]"}],"fun_headline_variants":["Affine Barbasch-Vogan: one orbit, one cell per level","Single orbit and left cell dictate L_k(g) at each integer level","Two conjectures: associated variety and simple modules unified","Level-by-level prediction for affine vertex algebras","Barbasch-Vogan analog: exact associated variety and block description"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument stands on a case-by-case match between the nilpotent orbit assigned to the level $m$ and the affine Weyl group cell assigned to the same $m$, plus the belief that the non-nilpotent part of the associated variety is exactly the center of the smallest Levi subalgebra containing the orbit; if either identification fails, the two conjectures fall.","fun_headline_variants_meta":{"raw":{"variants":["Affine Barbasch-Vogan: one orbit, one cell per level","Single orbit and left cell dictate L_k(g) at each integer level","Two conjectures: associated variety and simple modules unified","Level-by-level prediction for affine vertex algebras","Barbasch-Vogan analog: exact associated variety and block description"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000272,"raw_usage":{"total_tokens":1603,"prompt_tokens":886,"completion_tokens":717,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":502,"completion_tokens_details":{"reasoning_tokens":633}},"tokens_in":502,"tokens_out":717,"duration_ms":6441,"temperature":1.0,"reasoning_tokens":633,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T15:20:21.768114+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute $X_{L_k(\\mathfrak g)}$ for a simply-laced type not in the paper's tables, for example type $E_6$ at $m=3$ ($k=-9$), and compare with $\\operatorname{Ad}G\\cdot(d_L O_{\\check L}\\times\\mathfrak z(\\mathfrak l)\\times\\mathfrak u)$; any discrepancy falsifies Conjecture 5.1.1. Independently, list $\\mathrm{cl}_n$ on all nilpotent orbits of $E_8$ and check the equality $\\check O(m)=\\mathbf c(w_m)$ through Lusztig's cell bijection; one mismatch would invalidate Theorem 4.2.2 and with it both conjectures.","supporting_citations":[],"review_version":2}