{"id":"33cddfdd-24f5-4df4-909e-86c926e479d9","arxiv_id":"1908.03818","paper_version":5,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The author determines the full composition series and a canonical intertwining filtration of certain induced representations on p-adic symplectic and orthogonal groups, showing they are multiplicity-free of length 3^k.","lead":"This paper computes the exact composition series of a class of parabolically induced representations on p-adic classical groups that appear in the Mœglin-Tadić classification of discrete series. The result decomposes such representations into 3^k irreducible pieces via a canonical filtration, with applications to Jacquet modules and automorphic forms.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The paper's advertised 2^k-subrepresentation and 3^k-length counts fail in the one-half reducibility case (b_i = -1/2), which C1 explicitly allows.","rationale":"The reader identified Proposition 2.6's unverified reduction as the weakest assumption, but the more decisive problem is internal: the paper's own C1 conditions admit segments with b_i = -1/2, and for those segments Proposition 3.4 gives a composition series with fewer terms than Theorem 1.1 and Corollary 4.4 advertise. This is not a question of external scope or a missing proof; the advertised length 3^k and subrepresentation count 2^k are arithmetically inconsistent with the lemmas used to prove the main theorem. The core recursive formula (4.1) may be salvageable if the counts are corrected, but as written the central claimed consequence is false. The reader's concern about Proposition 2.6 is legitimate but secondary, so I disagree with the reader's identification of the weakest assumption.","tokens_in":16170,"tokens_out":18638,"duration_ms":184833,"concrete_test":"Take k=1 and choose ρ with Jord_ρ=∅ and ν^{1/2}ρ ⋊ σ_cusp reducible; set Δ = [ν^{1/2}ρ, ν^{3/2}ρ]. This satisfies C1 with b=-1/2 and satisfies C2 vacuously. Compute the composition series of δ(Δ)⋊σ using the paper's Proposition 3.4: it has two irreducible factors and one discrete-series subrepresentation. Theorem 1.1 predicts three irreducible factors and two subrepresentations. The contradiction is resolved only by restricting Theorem 1.1 and Corollary 4.4 to b_i ≠ -1/2 or by correcting the stated length and subrepresentation counts.","verdict_should_be":"REJECT","load_bearing_attack":"Under C1 (Section 3, first bullet), a segment Δ_i = [ν^{-b_i}ρ_i, ν^{c_i}ρ_i] with b_i = -1/2 is allowed whenever Jord_{ρ_i}=∅ and ν^{1/2}ρ_i ⋊ σ_cusp reduces. For such a segment, Proposition 3.4 (first case) gives δ(Δ_i) ⋊ σ = σ2 + Lang(δ(Δ_i) ⋊ σ), i.e. exactly one irreducible subrepresentation and total length 2. Proposition 3.3 and Proposition 3.5 likewise count discrete-series subrepresentations as 2^l, where l = #{i : -b_i ≠ 1/2}, not 2^k. If r of the k segments have b_i = -1/2, the total number of irreducible summands in the decomposition (4.1) is Σ_{X⊆S} 2^{|{j∉X : -b_j≠1/2}|} = 2^r 3^{k-r}, which is not 3^k when r>0. In particular Theorem 1.1 and Corollary 4.4, which assert 2^k irreducible subrepresentations and length 3^k without excluding b_i = -1/2, are false as stated. This is an internal inconsistency, not merely an unexpanded step: the paper's own Proposition 3.4 contradicts the advertised counts.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies parabolically induced representations of classical p-adic groups of the form δ(Δ_1)×⋯×δ(Δ_k)⋊σ, where σ is a discrete series obtained by extending a strongly positive discrete series σ_sp by a family of segments satisfying conditions (C1) and (C2). Condition (C1) restricts the exponents and Jordan-block data of each segment; condition (C2) requires pairwise irreducibility of the GL-inducements δ(Δ_i)×δ(Δ_j) and δ(~Δ_i)×δ(Δ_j). The main theorem (Theorem 4.1) asserts a multiplicity-one decomposition in the Grothendieck group into Langlands quotients, indexed by subsets X⊆{1,…,k} and by discrete-series subrepresentations σ′ of the remaining product, together with a filtration whose successive quotients are the displayed Langlands quotients. The proof is an induction combining the Mœglin-Tadić classification of discrete series, Tadić's Jacquet-module formula, and results of Muić on generalized principal series. The introduction states a special case (Theorem 1.1) in which the number of irreducible subrepresentations is 2^k and the total length is 3^k.","tokens_in":16414,"tokens_out":25121,"duration_ms":247887,"significance":"If valid, the result gives an explicit composition series for a substantial class of standard representations appearing in the Mœglin-Tadić classification, and the filtration via intertwining operators is a potentially useful tool for analysing Jacquet modules of discrete series. The proof is a serious induction based on the Mœglin-Tadić classification and Tadić's formula, and the main decomposition formula (4.1) is precise and checkable. The paper would be a useful contribution to the representation theory of p-adic classical groups, provided the scope of the counting claims and a few proof steps are clarified.","major_comments":[{"comment":"The family in (C1) explicitly allows b_i=-1/2 (equivalently -b_i=1/2). For such a segment, Proposition 3.4 gives δ(Δ_i)⋊σ = σ2 + Lang(δ(Δ_i)⋊σ), so the length is 2, not 3, and Proposition 3.5 counts only 2^l irreducible subrepresentations with l=#{i: -b_i≠1/2}. Substituting this into the sum in (4.1) gives total length 2^r 3^{k-r}, where r=#{i: -b_i=1/2}, not 3^k. Theorem 1.1 and Corollary 4.4 assert 2^k irreducible subrepresentations and, in Theorem 1.1, length 3^k without excluding b_i=-1/2. If these statements are intended only for segments produced by the Mœglin-Tadić reduction steps of Proposition 2.6, where the lower exponents are non-positive, that restriction must be stated explicitly; if they are intended for all families satisfying (C1)-(C2), they are false as written.","section":"Section 3, (C1), and Theorem 1.1 / Corollary 4.4"},{"comment":"The proof contains the step 'It is not hard to check that condition (C1) remained valid' after replacing linked segments by unions/intersections and possibly taking contragredients. No argument is supplied that the new endpoints satisfy the parity, emptiness, and cuspidal-reducibility conditions in (C1) relative to σ_sp, nor that the algorithm terminates with the required (C2) irreducibility. Since Proposition 2.6 is cited in Corollary 4.4 as the link between the Mœglin-Tadić description and the hypotheses of Theorem 4.1, this is a load-bearing gap. Please provide a complete verification or explicitly restrict Corollary 4.4 to families that already satisfy (C1)-(C2).","section":"Proposition 2.6"},{"comment":"The induction in the first case is described as an induction over card(Y), but the reducibility of δ([ν^{1/2}ρ,ν^cρ])⋊σ concerns the particular cuspidal ρ of the new segment. If no segment in Y has the same ρ, the 'minimal corresponding segment' [ν^{-b_j}ρ,ν^{c_j}ρ] is not defined and the reduction that removes the corresponding Jordan blocks from σ makes no reference to a ρ-segment. The argument should be formulated as an induction over the number of Y-segments with the given ρ (or should explain why segments with different ρ are irrelevant under (C2)). Without this clarification, the proof of the basic one-segment step is incomplete.","section":"Proposition 3.4, first case"}],"minor_comments":[{"comment":"The phrase 'there are 2k of them' should read 'there are 2^k of them' both in Theorem 1.1 and in Corollary 4.4; the superscript is essential to the counting claim.","section":"Theorem 1.1 and Corollary 4.4"},{"comment":"The sentence defining σ3 and σ4 says 'Jord(σ2)=Jord(σ3)=...'; it should say 'Jord(σ3)=Jord(σ4)=...'.","section":"Proposition 3.4, second case"},{"comment":"In the display after (4.7), 'j∈X\\S' should be 'j∈S\\X'; the current notation is nonsensical because X is a subset of S.","section":"Theorem 4.1 proof, equation (4.7)"},{"comment":"The summation condition 'card(X)=k' in (4.9) should be 'card(X)=l', consistent with the filtration index l in the statement.","section":"Corollary 4.3, equation (4.9)"},{"comment":"The reference 'By Remark 3.2 and Proposition 4.2 of [5]' is problematic: the paper has only Remark 3.1, no Remark 3.2. The intended reference should be corrected.","section":"Proposition 3.4, proof"},{"comment":"The notation 'Jord\\{(a,ρ)(a−,ρ)}' should be 'Jord\\{(a,ρ),(a−,ρ)}' with a comma separating the two pairs.","section":"Theorem 2.2(iii)"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper gives a full Grothendieck decomposition and a filtration for a class of induced representations built from discrete series, under the pairwise irreducibility condition (C2). That is a genuine advance over earlier work, which handled only cuspidal sigma or at most two segments, and the proof is a real induction using the Moeglin-Tadic classification and Tadic's Jacquet-module formula. The main theorem (4.1) and the filtration (4.2) appear structurally sound.\n\nThe catch: the advertised counts in Theorem 1.1 and Corollary 4.4 are wrong. Condition (C1) explicitly allows segments with b_i = -1/2 (first bullet, with -1/2 <= b_i). For such a segment, Proposition 3.4 (first case) gives exactly one irreducible subrepresentation and total length 2 for that step, not 3. Consequently the number of sigma' in the inner sum of (4.1) is 2^{l(S\\X)} where l counts indices with -b_i != 1/2, and the total length is 2^r 3^{k-r}, where r is the number of half-integral b_i. That is not 3^k when r>0. The claim of 2^k irreducible subrepresentations in Theorem 1.1 and Corollary 4.4 is similarly false unless r=0. The stress-test note is correct: this is an internal inconsistency with the paper's own Proposition 3.4.\n\nThe repair is local. The decomposition formula (4.1) itself can be stated with the correct multiplicities, and the induction proof seems to go through with those corrected counts. The paper should also either exclude b_i = -1/2 from the statements of Theorem 1.1 and Corollary 4.4 or state the corrected 2^a 3^b counts.\n\nTwo further soft spots, in proportion: Proposition 2.6's reduction to condition (C2) is asserted with \"It is not hard to check\" and deserves a detailed proof, since the actual scope of the theorem depends on it. And there are minor typos, e.g. in Proposition 3.4 the text writes Jord(sigma2) where it should be Jord(sigma3) and Jord(sigma4). None of this affects the core induction.\n\nFor a specialist in p-adic representation theory, this is a usable and valuable result once the counts are corrected. It is not ready as is because of the false statements, but the flaw is limited and fixable. Yes, send it to peer review; a competent referee should catch this and ask for revision.","headline":"The main composition-series formula looks right, but the advertised counts 2^k and 3^k are false when any segment has b_i=-1/2, a case the paper explicitly allows; the fix is to restate the counts with 2^a 3^b.","tokens_in":16993,"tokens_out":5377,"would_cite":true,"duration_ms":53104,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["22D30","22E50","22D12","11F85"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper establishes an explicit multiplicity-free composition series, with $3^k$ terms indexed by subsets of the segments, for a class of induced representations built on discrete series of classical p-adic groups.","keywords":["composition series","discrete series","induced representations","classical p-adic groups","Mœglin–Tadić classification","Langlands quotients","multiplicity-free representations","Jacquet modules"],"falsifier":"A direct calculation for $k=2$ with one explicit segment family satisfying (C1)–(C2) would settle the claim: use the Tadić formula (2.1) to compute the semisimplification of $\\delta(\\Delta_1)\\times\\delta(\\Delta_2)\\rtimes\\sigma$ and check that it has exactly nine irreducible subquotients, each with multiplicity one and the predicted Langlands quotient; a different length or a repeated quotient would disprove Theorem 4.1.","tokens_in":15908,"feed_emoji":"🧩","tokens_out":12262,"duration_ms":114199,"temperature":0.7,"pith_summary":"This paper determines the full composition series for a class of induced representations that arises inside the Mœglin–Tadić classification of discrete series for classical $p$-adic groups. The author proves that, under a pairwise irreducibility condition on the segments involved, such a representation is multiplicity-free and splits into exactly $3^k$ irreducible subquotients, indexed by subsets of the $k$ segments; each subquotient is an explicitly named Langlands quotient. The result matters because these induced representations are the building blocks out of which discrete series and standard representations are constructed, so an explicit decomposition feeds directly into the study of Jacquet modules and of automorphic forms. The proof works by an intertwining-operator filtration and a counting argument in the Grothendieck group.","feed_headline":"Discrete series inductions split into 3^k subquotients","feed_subtitle":"A full composition-series formula for the building blocks of discrete series in classical p-adic groups.","key_machinery":"The engine of the proof is a family of intertwining operators that permute the segments $\\Delta_i$ and eventually replace them by their contragredient segments $\\widetilde{\\Delta}_i$. For each $l$, the image $V_l$ of the sum of intertwinings from the representations attached to subsets $X$ of size $l$ gives a filtration $\\{0\\}=V_{-1}\\subseteq V_0\\subseteq\\cdots\\subseteq V_k=\\prod_{i\\in S}\\delta(\\Delta_i)\\rtimes\\sigma$, and Theorem 4.1 identifies each quotient $V_l/V_{l-1}$ with the direct sum of the Langlands quotients indexed by subsets $X$ of size $l$. The irreducibility condition (C2) is what makes the segment permutations isomorphisms and keeps the counting of subrepresentations exact; the Tadić formula for Jacquet modules is then used to show each candidate Langlands quotient occurs with multiplicity one.","core_discovery":"The central claim is that a large class of induced representations of classical $p$-adic groups—those built by parabolically inducing a product of essentially square-integrable general-linear representations onto a discrete series representation—has a fully explicit, multiplicity-free composition series. If $\\sigma$ is a discrete series obtained by extending a strongly positive discrete series $\\sigma_{\\mathrm{sp}}$ along segments satisfying the conditions (C1) and (C2), then for any further such family $\\{\\Delta_i: i \\in S\\}$ the representation $\\prod_{i \\in S}\\delta(\\Delta_i)\\rtimes\\sigma$ has exactly $3^{|S|}$ irreducible subquotients. In the Grothendieck group it equals $$\\sum_{X \\subseteq S}\\sum_{\\$\\sigma$' \\hookrightarrow \\prod_{j \\in S\\setminus X}\\delta(\\Delta_j)\\rtimes\\$\\sigma$}\\mathrm{Lang}\\Bigl(\\prod_{i \\in X}\\delta(\\Delta_i)\\rtimes\\$\\sigma$'\\Bigr),$$ and these Langlands quotients occur as the successive quotients of an explicit filtration by images of intertwining operators. The paper also derives the decomposition of the same inductions when the starting piece is a Langlands quotient rather than a discrete series.","pith_inferences":["Beyond the paper, filling in the sketched reduction in Proposition 2.6 would extend the same $3^k$ formula to every discrete series from the Mœglin–Tadić classification, including those whose defining segments are linked.","Beyond the paper, because the proof replaces linked segments by their union and intersection, the general linked case is likely describable by an inclusion–exclusion over segment intersections, with additional terms for each reducible pair.","Beyond the paper, the method appears transferable to settings with an analogous discrete-series classification and Jacquet-module formula, such as metaplectic groups, though the paper itself treats only symplectic and orthogonal groups.","Beyond the paper, a small-rank numerical test of Corollary 4.3 for $k=2$ would identify any boundary cases where the irreducibility assumption is doing more work than the proof makes visible."],"forward_implications":["In the irreducible case the standard representation $\\delta(\\Delta_1)\\times\\cdots\\times\\delta(\\Delta_k)\\rtimes\\sigma_{\\mathrm{sp}}$ has length $3^k$, so its semisimplification can be written down directly from the formula without recursion.","Every irreducible subrepresentation of the induction is a discrete series extension of $\\sigma_{\\mathrm{sp}}$, and the theorem specifies exactly which Jordan blocks are added and which values of the $\\epsilon$-function occur.","The same filtration works when the base is a Langlands quotient rather than a discrete series (Corollary 4.3), extending the decomposition to a broader class of induced representations.","Combined with known Jacquet-module formulas for strongly positive discrete series, the decomposition gives a direct route to the Jacquet modules of a large family of discrete series.","The multiplicity-one property and the exact $3^k$ count give a concrete numerical check for any attempt to decompose these representations algorithmically."],"supporting_citations":[{"why":"Supplies the base case: composition series of generalized principal series attached to strongly positive discrete series, used as Proposition 3.4 and for constructing discrete series extensions.","marker":"[7]"},{"why":"Provides the Mœglin–Tadić construction of discrete series as subrepresentations of inductions of the type being decomposed.","marker":"[6]"},{"why":"Gives the Tadić formula for Jacquet modules, used to count occurrences of each candidate subquotient.","marker":"[9]"},{"why":"Establishes the segment notation and the irreducibility criterion for general-linear inductions that defines $\\delta(\\Delta)$ and condition (C2).","marker":"[12]"},{"why":"Supplies the classification of discrete series by admissible triples, used to describe extensions of $\\sigma_{\\mathrm{sp}}$.","marker":"[5]"},{"why":"Computes Jacquet modules of strongly positive discrete series, used in Proposition 2.5 to constrain cuspidal supports.","marker":"[4]"},{"why":"Prior treatment of the same irreducibility condition with a cuspidal starting point, which the paper extends to general discrete series.","marker":"[10]"},{"why":"Solves the two-segment cuspidal case, serving as a special case and a check on the general formula.","marker":"[2]"}],"fun_headline_variants":["3^k subquotients: explicit composition series for p-adic inductions","Induced reps over p-adic groups split into 3^|S| pieces","Exact subquotient count: 3^|S| for induced discrete series","Counting 3^|S| subquotients in induced representations"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument stands or falls on the irreducibility condition (C2) for every pair of segments—both $\\delta(\\Delta_i)\\times\\delta(\\Delta_j)$ and $\\delta(\\widetilde{\\Delta}_i)\\times\\delta(\\Delta_j)$ irreducible for $i\\neq j$—together with the reduction in Proposition 2.6 asserting that every discrete series can be brought into this situation, a step the paper sketches rather than proves in detail.","fun_headline_variants_meta":{"raw":{"variants":["3^k subquotients: explicit composition series for p-adic inductions","Induced reps over p-adic groups split into 3^|S| pieces","Exact subquotient count: 3^|S| for induced discrete series","Counting 3^|S| subquotients in induced representations"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000988,"raw_usage":{"total_tokens":4138,"prompt_tokens":842,"completion_tokens":3296,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":458,"completion_tokens_details":{"reasoning_tokens":3211}},"tokens_in":458,"tokens_out":3296,"duration_ms":21954,"temperature":1.0,"reasoning_tokens":3211,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:00:22.511155+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct calculation for $k=2$ with one explicit segment family satisfying (C1)–(C2) would settle the claim: use the Tadić formula (2.1) to compute the semisimplification of $\\delta(\\Delta_1)\\times\\delta(\\Delta_2)\\rtimes\\sigma$ and check that it has exactly nine irreducible subquotients, each with multiplicity one and the predicted Langlands quotient; a different length or a repeated quotient would disprove Theorem 4.1.","supporting_citations":[{"cited_title":"Mui´ c, Composition series of generalized principal series; the ca se of strongly positive dis- crete series , Israel J","cited_arxiv_id":null,"evidence_quote":"Supplies the base case: composition series of generalized principal series attached to strongly positive discrete series, used as Proposition 3.4 and for constructing discrete series extensions."},{"cited_title":"Tadi´ c,Construction of discrete series for classical p-adic groups, J","cited_arxiv_id":null,"evidence_quote":"Provides the Mœglin–Tadić construction of discrete series as subrepresentations of inductions of the type being decomposed."},{"cited_title":"Tadi´ c,Structure arising from induction and Jacquet modules of rep resentations of classical p-adic groups , J","cited_arxiv_id":null,"evidence_quote":"Gives the Tadić formula for Jacquet modules, used to count occurrences of each candidate subquotient."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the segment notation and the irreducibility criterion for general-linear inductions that defines $\\delta(\\Delta)$ and condition (C2)."},{"cited_title":"Mœglin, Sur la classification des s´ eries discr` etes des groupes cla ssiques p-adiques: param` etres de Langlands et exhaustivit´ e","cited_arxiv_id":null,"evidence_quote":"Supplies the classification of discrete series by admissible triples, used to describe extensions of $\\sigma_{\\mathrm{sp}}$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Computes Jacquet modules of strongly positive discrete series, used in Proposition 2.5 to constrain cuspidal supports."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Prior treatment of the same irreducibility condition with a cuspidal starting point, which the paper extends to general discrete series."},{"cited_title":"Ciganovi´ c, Composition series of a class of induced representations, a case of one half cuspidal reducibility, Pacific J","cited_arxiv_id":null,"evidence_quote":"Solves the two-segment cuspidal case, serving as a special case and a check on the general formula."}],"review_version":1}