{"id":"88c45f0d-0311-48a3-892e-db036bb3b6b9","arxiv_id":"2501.09085","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper constructs and proves the Vogan reduction map from irreducible SL_n(k)-representations to tame Langlands parameter inertia classes for SL_n(F), with fiber and compatibility theorems.","lead":"This paper proves Vogan's conjecture for the special linear group: irreducible representations of SL_n over a finite field are matched with inertia classes of tame Langlands parameters for SL_n over a p-adic field with the same residue field. It is the first such complete result beyond the general linear group and connects finite group representations to the local Langlands correspondence.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Surjectivity of eta_* in Lemma 3.3 rests entirely on a cited lifting theorem; if [10] does not cover semisimple tame projective Weil representations, the main surjection may be missing classes.","rationale":"I read the paper as a coherent extension of the GL_N results to SL_N. The proofs of Sections 4 and 5 are largely internal, and I found no internal contradiction in the constructions of the Macdonald-Vogan correspondence, the fiber parametrization, or the compatibility theorems. The single place where a non-obvious external input is used without proof is the surjectivity of eta_* in Lemma 3.3. The reader's weakest_assumption pinpoints exactly this. My recommendation is to accept the paper conditional on a verification of the lifting theorem for the tame Weil quotient, since the central surjectivity claim logically hinges on it. If the check confirms that every tame PGL_N(C)-valued Weil representation lifts to a semisimple GL_N(C)-valued representation, Lemma 3.3 is sound and the paper's claims stand. If not, the map M^1_N may miss entire inertia-equivalence classes, invalidating the main theorem. This is a concrete, checkable point rather than a stylistic concern, and it does not affect the reader's overall assessment unless the lifting theorem fails. Therefore I recommend CONDITIONAL acceptance pending that verification.","tokens_in":28903,"tokens_out":19894,"duration_ms":189449,"concrete_test":"Independently compute H^2_cont(W_F/P_F, C^*) via the Hochschild-Serre spectral sequence for Gamma = Z ⋉ T, with T the prime-to-p procyclic inertial quotient on which Frobenius acts by q. If H^2 = 0, construct for a sample tame parameter (e.g., N = 2, inertia image of order m, Fr a nontrivial normalizing element) an explicit GL_2(C) lift. If H^2 ≠ 0, exhibit the corresponding projective parameter and test whether it lifts. This settles whether the surjectivity step of Lemma 3.3 holds.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Lemma 3.3 defines eta_* as the bridge from GL_N IF-classes modulo k_F^* to SL_N IF-classes. Its surjectivity is proved in the final paragraph by citing [10] for the existence of a semisimple lift rho in (Phi_N)^0 of every tame projective Weil representation rho-bar in (Phi'^1_N)^0. This is the only place where the paper goes from PGL_N(C)-valued parameters to GL_N(C)-valued ones, so if the cited result does not produce a lift with rho|_P_F = 1 and rho(Fr) semisimple, the map M^1_N in (23) may not be surjective and the main theorem fails for the missing classes. The paper gives no statement of the theorem in [10], no verification of its hypotheses, and no fallback construction. The cited paper predates the modern LLC for GL_N and is about conducteurs and epsilon factors; whether it contains exactly this lifting statement is not self-evident. Although the statement is likely true, because the tame Weil quotient W_F/P_F is a semidirect product of Z by a procyclic prime-to-p group whose Schur multiplier may vanish, the argument as written is a one-sentence citation at a load-bearing point.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper proves Vogan's conjecture for SL_N in the tame/depth-zero setting. The author defines a Macdonald–Vogan correspondence M^1_N from irreducible representations of the finite group SL_N(k_F) to IF-equivalence classes of tame Langlands parameters for SL_N(F); proves in Theorem 4.3 that each fiber is a torsor for the character group of the relevant component group; and proves Theorems 5.5 and 5.15 giving compatibility with the Gelbart–Knapp local Langlands correspondence for SL_N, including compatibility of the component-group parametrizations. The proof proceeds by reducing to GL_N via Clifford theory and the bijection η_* between GL_N and SL_N parameters.","tokens_in":29163,"tokens_out":9921,"duration_ms":110443,"significance":"This is a meaningful step beyond the GL_N case and gives the first explicit verification of Vogan's conjecture for a group whose dual has nontrivial component groups in the tame setting. The paper combines Macdonald's correspondence, the Gelbart–Knapp LLC for SL_N, Schneider–Zink heads of parahoric restriction, and multiplicity-free restriction in a careful way. The examples in Section 4.1 are illuminating, and the main compatibility theorems are proved in detail. The central caveat is that surjectivity of the bridge η_* is delegated to a one-sentence citation; once that input is made explicit, the central claims of the paper appear sound.","major_comments":[{"comment":"The surjectivity of η_* — and hence of M^1_N in (23) — rests entirely on the final sentence of the lemma, which states that the existence of a semisimple Weil representation lift was proved in [10]. The manuscript does not state the precise theorem from [10], nor does it verify that the lift satisfies the three conditions needed for a tame Langlands parameter in (Φ_N)^0: triviality on P_F, semisimplicity of ρ(Fr), and the compatibility condition on E. For example, a projective lift could in principle be scalar on P_F and would then have to be adjusted by a character. Please include the precise statement of the cited result, verify its hypotheses, or give a short self-contained proof of this lifting fact.","section":"§3.1, Lemma 3.3"}],"minor_comments":[{"comment":"Condition (a') contains a typo: it should read Aρ1|_{I_F}A^{-1} = ρ2|_{I_F}, not Aρ1|_{I_F}A^{-1} = ρ2.","section":"§3.1, Definition 3.1"},{"comment":"The symbols 'ä yF^*', 'ä z WF' and 'ä xk_F^*' appear to be typesetting artifacts; the quotient groups in these diagrams should be displayed with standard notation for the orbit spaces.","section":"§5.1, diagrams (46) and (47)"},{"comment":"The equality CPGL_N(1, PGL_N(C), 0) = PGL_N(C) depends crucially on the convention that C_H(X) for a subset X is the stabilizer of X under conjugation, not the pointwise centralizer. This convention is correct but easy to misread; a brief reminder at that point would improve clarity.","section":"§4.1, Example 1"},{"comment":"The notation HPG1_N G1_N is used for the head of parahoric restriction of a single representation as well as for a sum over a coset of the component-group character group; the two uses could be distinguished notationally.","section":"§5.2, Definition 5.13"}],"recommendation":"major_revision","confidential_remarks":"The only substantive issue is the unstated lifting theorem in Lemma 3.3. The rest of the paper is carefully argued and the central theorems are plausible. This is a fixable issue rather than a fatal one: the author should either prove the lifting lemma or quote [10] with a precise statement of the theorem and a verification of its hypotheses. There is no concern about fit with the journal's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Vogan's conjecture for SL_n is the genuine result here, and I think it holds together. Collacciani constructs a surjection from irreducible reps of SL_n(k) to inertia-equivalence classes of tame Langlands parameters for SL_n(F), parametrizes the fibers by component groups, and proves compatibility with the LLC for SL_n, extending the GL_n results of Macdonald and Silberger-Zink. That is a real theorem, not a repackaging: the disconnected center of PGL_n(C) creates genuine complications, and she deals with them head-on, including the noncanonical nature of the fiber parametrization and the fact that the natural map between the two component groups is neither injective nor surjective. The examples in Section 4.1 are useful and honest.\n\nThe paper is well-organized. It leans heavily on established results—Gelbart-Knapp, the LLC for GL_n, Macdonald, Schneider-Zink—but that is appropriate. The proofs I checked appear sound; I found no circularity and no back-fitting. The central construction M^1_N is defined independently of the LLC for SL_n, and compatibility is proved, not assumed.\n\nSoft spot: the surjectivity of eta_* in Lemma 3.3 rests entirely on a cited theorem from [10] (Henniart, 1977-78) that every tame projective Weil representation lifts to a semisimple GL_N(C)-valued representation. This is the one load-bearing external input that is not stated precisely, and if it fails, the main surjection loses classes. I would not bet against it—the tame Weil quotient is semidirect and the Schur multiplier should behave—but the paper should quote the theorem or give an argument. The referee should check this specific point. Minor issues: some typos and heavy notation, but not a barrier.\n\nThis paper is for specialists in p-adic and finite group representation theory, particularly anyone working on Vogan's program. It deserves a serious referee. I recommend sending it to review, with the lifting theorem as the point to verify.","headline":"A solid new theorem: Vogan's conjecture for SL_n is proved, with one load-bearing cited lift that should be checked.","tokens_in":29665,"tokens_out":2164,"would_cite":true,"duration_ms":22699,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["22E50","20C33","22E35"],"pacs":[],"model":"deepseek-v4-flash","headline":"For SL_n, irreducible representations of SL_n(k) surject onto inertia-equivalence classes of tame Langlands parameters for SL_n(F), with fibers parametrized by component-group characters, compatible with the local Langlands correspondence.","keywords":["tame local Langlands correspondence","SL_n","finite groups of Lie type","Macdonald correspondence","Vogan conjecture","parahoric restriction","component groups","inertia equivalence"],"falsifier":"Find a tame Langlands parameter for SL_N(F) whose projective Weil representation admits no semisimple lift to GL_N(C); the paper relies on [10] for the existence of all such lifts. A more concrete test: compute the fiber size of $M^{1}$_N over a parameter and compare it with the order of the component group; mismatches would signal an error in the parametrization.","tokens_in":2060,"feed_emoji":"🧮","tokens_out":4387,"duration_ms":115106,"temperature":0.7,"pith_summary":"Vogan's conjecture says that the irreducible representations of a finite group of Lie type should admit a Langlands-style parametrization by inertia classes of tame Langlands parameters of the corresponding p-adic group, with fibers indexed by characters of component groups. This paper proves that conjecture for SL_n. It constructs the Macdonald–Vogan correspondence $M^{1}$_N, a surjection from the set of irreducible representations of SL_n(k) (k the residue field of F) to the inertia-equivalence classes of tame Langlands parameters for SL_n(F), and proves that each fiber is a torsor for the character group of an explicitly defined component group. The main compatibility theorems (Theorems 5.5 and 5.15) show that this finite-field packet structure matches the depth-zero local Langlands correspondence for SL_n(F): the head of parahoric restriction of a depth-zero representation lands in the fiber of its parameter's inertia class, and the component-group labels agree through the natural map. This matters because it extends to SL_n a structural picture previously available only for GL_n, where centralizers are connected.","feed_headline":"SL_n finite-group packets match tame Langlands classes","feed_subtitle":"Representations of SL_n(k) split by component-group characters, matching depth-zero p-adic packets.","key_machinery":"The central object is the Macdonald–Vogan correspondence $M^1_N$, built from the classical Macdonald correspondence for $GL_N$ and Clifford theory. The distinguishing feature of the $SL_N$ case is that an inertia-equivalence class must also record the coset $\\rho(\\mathrm{Fr}) C^0_{\\mathrm{PGL}_N(\\mathbb C)}(\\rho|_{I_F})$ of the Frobenius image in the connected centralizer of the inertia image; this extra data is what allows the component group of the full stabilizer to appear. The other main ingredient is the head of parahoric restriction $HP$: it selects the multiplicity-one top constituent of the reduction of a depth-zero representation, and its equivariance properties (Propositions 5.10 and 5.11) are what make the compatibility theorem work.","core_discovery":"The central claim is that Vogan's conjecture holds for $SL_N$. Concretely, the paper constructs a surjection $M^1_N : \\Omega^1_N \\to (\\Phi^1_N)_0/\\sim_{I_F}$ from irreducible representations of $SL_N(k)$ to inertia-equivalence classes of tame Langlands parameters for $SL_N(F)$ (Definition 3.4 and Theorem 4.3). Each fiber carries a simply transitive action of the character group of the component group $A_{\\mathrm{PGL}_N(\\mathbb C)}(\\rho|_{I_F}, \\rho(\\mathrm{Fr}) C^0_{\\mathrm{PGL}_N(\\mathbb C)}(\\rho|_{I_F}), E)$. The paper then proves Theorems 5.5 and 5.15, which state that this finite-field packet structure is compatible with the depth-zero local Langlands correspondence for $SL_N(F)$: if $\\pi_{\\rho,E}\\in L^1_N{}^{-1}(\\rho,E)$ and $\\pi_{\\rho,E}\\in M^1_N{}^{-1}((\\rho,E)_{I_F})$ correspond under the head of parahoric restriction, then the component-group labels are related by the natural map $\\iota$. In particular, $M^1_N$ is not merely analogous to the Langlands correspondence but is its finite-field shadow.","pith_inferences":["The same reduction might apply to other split reductive groups, but the component groups are generally nonabelian, so fiber parametrization would be more intricate.","The compatibility suggests that the head of parahoric restriction is the natural functor between the depth-zero Langlands correspondence and the finite-field packet correspondence, and could be studied for other groups.","A computational check for small N and q could enumerate tame parameters and finite-group representations to test the fiber sizes and the lifting assumption of Lemma 3.3."],"forward_implications":["For each tame parameter, the fiber of the Macdonald–Vogan map is a torsor for the appropriate component-group character group, so choosing a base point gives a canonical bijection between finite-group representations and characters.","The depth-zero local Langlands correspondence for SL_n(F) is compatible with the finite-field parametrization, so this gives a way to compute L-packets by parahoric restriction.","The compatibility persists when the maximal compact subgroup is changed, up to conjugating the base representation.","The fibers of the two correspondences can have different sizes; the paper gives explicit examples where one is larger than the other, and vice versa.","The action of the component-group character on L-packets maps via the natural map to the action on Macdonald–Vogan fibers, so labels transform in a controlled way."],"supporting_citations":[{"why":"States the conjecture that finite groups of Lie type admit a Langlands-style parametrization via inertia classes and component groups; the paper's goal is to prove this for SL_n.","marker":"[22]"},{"why":"Constructs the Macdonald correspondence for GL_n, the finite-field parametrization by I_F-equivalence classes that M^1_N builds on.","marker":"[17]"},{"why":"Proves compatibility of Macdonald's correspondence with the depth-zero local Langlands correspondence for GL_n; the paper reformulates and extends this to SL_n.","marker":"[20]"},{"why":"Provides the local Langlands correspondence for SL_N(F) and the simple transitive action of component-group characters on L-packets, used throughout Sections 2 and 5.","marker":"[8]"},{"why":"Supplies the lifting theorem for projective Weil representations to GL_N(C); the surjectivity of eta_* in Lemma 3.3 depends on it.","marker":"[10]"},{"why":"Gives the lemma identifying K^+_N-fixed points with K^{1+}_N-fixed points, used to compare parahoric restriction for GL_N and SL_N.","marker":"[7]"},{"why":"Its Proposition 6.2 underlies Theorem 1.1 describing the head of parahoric restriction as the top constituent of the reduction.","marker":"[18]"}],"fun_headline_variants":["Vogan's conjecture proven for SL_n","Surjection links SL_n reps to tame Langlands classes","Depth-zero LLC matches finite-field packets for SL_n","SL_n finite-field shadow of Langlands correspondence"],"cache_read_input_tokens":31872,"weakest_assumption_plain":"The argument needs every tame projective Weil representation of the Weil group into PGL_N(C) to be liftable to a semisimple representation into GL_N(C); if any such parameter lacks a lift, surjectivity of the map eta_* and hence of the whole correspondence fails.","fun_headline_variants_meta":{"raw":{"variants":["Vogan's conjecture proven for SL_n","Surjection links SL_n reps to tame Langlands classes","Depth-zero LLC matches finite-field packets for SL_n","SL_n finite-field shadow of Langlands correspondence"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000765,"raw_usage":{"total_tokens":3373,"prompt_tokens":904,"completion_tokens":2469,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":520,"completion_tokens_details":{"reasoning_tokens":2408}},"tokens_in":520,"tokens_out":2469,"duration_ms":17105,"temperature":1.0,"reasoning_tokens":2408,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T20:10:24.120152+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find a tame Langlands parameter for SL_N(F) whose projective Weil representation admits no semisimple lift to GL_N(C); the paper relies on [10] for the existence of all such lifts. A more concrete test: compute the fiber size of $M^{1}$_N over a parameter and compare it with the order of the component group; mismatches would signal an error in the parametrization.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"States the conjecture that finite groups of Lie type admit a Langlands-style parametrization via inertia classes and component groups; the paper's goal is to prove this for SL_n."},{"cited_title":"Macdonald","cited_arxiv_id":null,"evidence_quote":"Constructs the Macdonald correspondence for GL_n, the finite-field parametrization by I_F-equivalence classes that M^1_N builds on."},{"cited_title":"Silberger and Ernst-Wilhelm Zink","cited_arxiv_id":null,"evidence_quote":"Proves compatibility of Macdonald's correspondence with the depth-zero local Langlands correspondence for GL_n; the paper reformulates and extends this to SL_n."},{"cited_title":"Gelbart and Anthony W","cited_arxiv_id":null,"evidence_quote":"Provides the local Langlands correspondence for SL_N(F) and the simple transitive action of component-group characters on L-packets, used throughout Sections 2 and 5."},{"cited_title":"Représentations du groupe de weil d’un co rps local : conducteurs et facteurs ǫ","cited_arxiv_id":null,"evidence_quote":"Supplies the lifting theorem for projective Weil representations to GL_N(C); the surjectivity of eta_* in Lemma 3.3 depends on it."},{"cited_title":"Bushnell and Paul C","cited_arxiv_id":null,"evidence_quote":"Gives the lemma identifying K^+_N-fixed points with K^{1+}_N-fixed points, used to compare parahoric restriction for GL_N and SL_N."},{"cited_title":"K-types for the tempered components of a p-adic general linear group","cited_arxiv_id":null,"evidence_quote":"Its Proposition 6.2 underlies Theorem 1.1 describing the head of parahoric restriction as the top constituent of the reduction."}],"review_version":1}