{"id":"62b45cfe-7c96-4b76-95e9-70af04aa0841","arxiv_id":"1908.07833","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Every indecomposable trivial source module in a block with a cyclic defect group is explicitly parametrized by a path, direction, and multiplicity on the Brauer tree.","lead":"This paper fully classifies the indecomposable trivial source modules in p-blocks with cyclic defect groups, giving their exact location in the stable Auslander-Reiten quiver and their Brauer tree paths. It closes a classification question that earlier work had only answered partially.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 5.3's completeness inherits the new correction to the Hiss–Naehrig classification in Theorem A.1; the exceptional-leaf case is only sketched and needs independent verification.","rationale":"The reader's weakest_assumption identifies exactly the same load-bearing concern: the corrected Hiss–Naehrig classification in Appendix A, especially the exceptional-leaf case, is the external theorem whose completeness Theorem 5.3 inherits. I agree with that diagnosis. The paper flags the correction explicitly and gives a plausible explanation, but the proof is a brief sketch rather than a full derivation, and no independent check is provided. My stress-test did not uncover an actual error in the correction, the distance formulas, or the final classification; the concern is that a subtle mistake in the exceptional-leaf swap would directly invalidate the completeness of Theorem 5.3. Because the reader already weighed this uncertainty and chose ACCEPT with MODERATE confidence, and because I have not produced a concrete counterexample or a demonstrated flaw, I do not recommend moving the verdict. The proposed concrete test would settle the concern: if the corrected list in Theorem A.1(c) matches the set of liftable modules selected by the distance criterion in a small but non-trivial exceptional-leaf case, the completeness of Theorem 5.3 is supported. Until such a check is run, the residual risk is exactly the one the reader flagged, and the ACCEPT verdict with moderate confidence remains reasonable.","tokens_in":22994,"tokens_out":17302,"duration_ms":161501,"concrete_test":"Independently verify Theorem A.1(c) in the exceptional-leaf case. Concretely: take a cyclic block with e = 2, m = 2 (e.g., defect group C5 and a Brauer tree that is a star with exceptional leaf). Enumerate all indecomposable modules via Janusz paths, direction, and multiplicity, compute each module's distance to the boundary using [BC02, Theorem 3.5], and compare the subset satisfying the distance criterion of Theorem A.1(b) with the corrected list (1)–(7) of Theorem A.1(c). If the two sets coincide (count and parameters), the correction is supported; any mismatch, in particular in case (3)(ii) with χΛ a leaf, shows the classification in Theorem 5.3 is incomplete. Running this for p = 5, e = 2, m = 2 and for p = 7, e = 2, m = 3 would test both parity and length regimes.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The completeness of Theorem 5.3 is inherited verbatim from the corrected Hiss–Naehrig classification in Theorem A.1. Every trivial source module is liftable, so the 'if and only if' in Theorem 5.3(b) reduces to selecting, among the liftable modules listed in Theorem A.1(c), those whose distance to the positive boundary equals d_+(X) = ℓ_i p^{n-i} - 1. The paper introduces a genuine modification of the published theorem: when the exceptional vertex χΛ is a leaf of σ(B), Theorem A.1 adds case (c)(2') (the simple module labelling the leaf is not a hook) and alters case (c)(3) by swapping the roles of that simple module and the length-m hook with µ = m composition factors. The proof of this correction is a single informal paragraph ('the roles ... need to be swapped') rather than a full re-derivation. No independent verification or computational check is supplied. Since a missed liftable module in the exceptional-leaf case would propagate directly into Theorem 5.3, the central completeness claim is only as secure as this correction. The distance computations in Appendix B and the divisibility conditions in Theorem 5.3 all presuppose that Theorem A.1(c) is exhaustive and contains no erroneous multiplicity bounds (e.g., 2 ≤ µ ≤ m−1 in (3)(ii)).","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper gives a complete classification of the indecomposable trivial source modules in a p-block with non-trivial cyclic defect group of order p^n. The authors first determine, in Theorem 5.1, the distance of any such module to the positive boundary of the stable Auslander-Reiten quiver in terms of the integer ℓ_i obtained from the Dade-group restriction of the endo-permutation module W parametrising the source algebra of the block. They then combine this distance formula with a corrected version (Appendix A) of the Hiss-Naehrig classification of indecomposable liftable modules in cyclic blocks, together with distance computations (Appendix B), to obtain a full classification in Theorem 5.3: each trivial source module with vertex D_i is described by a path on the Brauer tree, a direction, and a multiplicity, subject to explicit congruence conditions and closed-form expressions for the multiplicity. The paper also treats the exceptional case where the exceptional vertex is a leaf of the Brauer tree, corrects the statement of [HN12], and includes the case D ≅ C_2 separately.","tokens_in":23274,"tokens_out":13970,"duration_ms":134156,"significance":"If the results are correct, this is a substantial contribution: it completes the classification of trivial source modules in blocks with cyclic defect groups, a question that had been addressed only partially by Michler, Bessenrodt, and Koshitani-Kunugi. The proof strategy is strong: the reduction from the block B to the nilpotent centraliser block c via Green correspondence and source algebra equivalence is coherent, and the distance formula is derived from the Dade group structure of the endo-permutation module rather than assumed. The paper is careful to correct and extend a prior classification of the same first author, and it provides an explicit appendix with distance computations. The claims are falsifiable: Corollary 5.2 gives concrete values for principal blocks and for vertex D_1, and the divisibility conditions in Theorem 5.3 can be checked in examples. The main caveats concern the completeness of the corrected liftable-module classification and the sufficiency of the distance criterion in the proof of Theorem 5.3.","major_comments":[{"comment":"The correction to the Hiss-Naehrig classification in the case where the exceptional vertex is a leaf of the Brauer tree is not actually proved. The paragraph after the statement merely says that the simple module labelling the leaf is not a hook and that 'the rôles ... need to be swapped in the original proof.' Since Theorem 5.3(b) inherits exhaustiveness directly from Theorem A.1(c), a missed or misdescribed liftable module in this leaf case would propagate into the main classification. Please provide a complete proof of the modified statement, or a precise reference to a proof, together with a justification that the count e(2m+1) of liftable modules is unchanged by the correction.","section":"Appendix A, Theorem A.1(c)(2') and (c)(3)"},{"comment":"The 'if' direction of the classification is not justified by the displayed distance equalities. In steps 3–8, the authors show that a liftable module X of a given path type has distance d(X,H) = ℓ_i p^{n-i} − 1 (or the equivalent negative-boundary condition), and then conclude 'by Theorem 5.1, X is a trivial source module'. But Theorem 5.1 only states a necessary condition: every trivial source module with vertex D_i has that distance. It does not state that any module with that distance is trivial source. A row of the stable AR-quiver can contain modules with different sources and vertices, so the distance equality alone does not identify the source as trivial. Please add an argument showing sufficiency—for example, by identifying the module explicitly as the Green correspondent of an induced module from the centraliser block (as in Lemma 4.6 and Corollary 4.5), or by proving a lemma that among the liftable modules of a given path type, the stated multiplicity condition characterises the modules whose trivial-source lift has non-negative character values on D_1.","section":"Theorem 5.3(b), proof steps 3–8"}],"minor_comments":[{"comment":"In the proof of parts (c), (d), and (e), the displayed formula for the even-l case reads e(µ−1)−1, while the statements and Theorem 5.3 use e(µ−1). Since (n−1)/2 + ηe = e + e(µ−2) = e(µ−1), the proof displays appear to contain arithmetic typos; please correct them.","section":"Appendix B, Proposition B.1(c)–(e)"},{"comment":"The index notation in the summation formula for ℓ_i is slightly ambiguous: the set {j | 0 ≤ i_j < i} is used before the indices i_j have been defined. For readability, define the sequence i_0 < i_1 < ... in the statement of Theorem 5.1 itself, rather than only in §4.5.","section":"§4.5 and Theorem 5.1"},{"comment":"The citation to Green's Theorem on Zeros of Characters is given as [CR81, (19.27)], but the proof would benefit from a short explanation of why the conjugates x_i x x_i^{-1} with x_i ∉ N_1 cannot lie in a subgroup that intersects D_1 trivially, rather than leaving this to the cited proof of [Alp86, §17, Theorem 3].","section":"Lemma 4.2"}],"recommendation":"major_revision","confidential_remarks":"The manuscript depends heavily on the authors' own previous classification [HN12], and the correction to that theorem is a load-bearing component of the main result. Given that the proof of the correction is only sketched, it would be advisable to have an independent verification of Theorem A.1(c)(2') and (c)(3) before publication. The paper otherwise fits the journal's scope and is a serious contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Punchline: this is a genuine completion of a long-running classification problem, not a repackaging. The distance formula in Theorem 5.1 and the full Brauer-tree-path classification in Theorem 5.3 are new, and the paper is honestly assembled from known ingredients. The main thing to check before relying on the completeness claim is the corrected Hiss–Naehrig classification in Appendix A.\n\nWhat is actually new: Michler, Bessenrodt, and Koshitani–Kunugi each had partial results; this paper gives exact locations in the stable AR-quiver and the full path/direction/multiplicity data for every trivial source module. The reduction to the centralizer block and the Dade group computation of ℓ_i are coherent. The distance formula is derived, not fitted. The proof of Theorem 5.1 flows cleanly through the source algebra, and the Green correspondence steps are standard.\n\nSoft spots: the completeness of Theorem 5.3 is inherited from the classification of liftable modules in Theorem A.1, and Theorem A.1 carries a correction to [HN12] for the case where the exceptional vertex is a leaf. The proof of that correction is one sentence—'the roles need to be swapped'—rather than a real derivation. Since every trivial source module is liftable, a missed liftable module would propagate directly into the main theorem. I think this is a fair referee request, not a demonstrated gap: the correction is explicit, the original proof is by the same first author, and the nearby statements are consistent. I would ask the authors to expand the argument or provide an independent check. There is also a small typo in Appendix B: in the proof of Proposition B.1(c)–(e), the even-l case displays e(µ−1)−1 while both the statement and the earlier computation give e(µ−1). The statement is what Theorem 5.3 uses, and it looks right. Citation pattern is clean: the self-citation to [HN12] is justified because the paper corrects and extends that result.\n\nBottom line: the paper settles the problem, the main derivation is sound, and the caveats are localized. I'd send it to a serious referee and expect acceptance after the Appendix A correction is firmed up. Worth bringing to reading group for anyone working on p-permutation modules or cyclic blocks.","headline":"Completes the classification of trivial source modules in cyclic defect blocks with explicit distance formulas; the only real caveat is the sketched correction to the Hiss–Naehrig classification in Appendix A.","tokens_in":23791,"tokens_out":4154,"would_cite":true,"duration_ms":37218,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["20C20"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that every non-projective indecomposable trivial source module in a block with a non-trivial cyclic defect group is determined by its Brauer-tree path, direction, and multiplicity, with distance fixed by an explicit…","keywords":["trivial source modules","p-permutation modules","blocks with cyclic defect groups","Brauer trees","liftable modules","stable Auslander-Reiten quiver","source algebras","endo-permutation modules"],"falsifier":"One concrete test: in a principal block with cyclic defect group $C_{p^n}$, the theorem predicts that every trivial source module with vertex $D_i$ lies exactly $p^{n-i}-1$ rows from the positive boundary of the stable Auslander-Reiten quiver. Compute that quiver explicitly and inspect the sources of the liftable modules at the predicted rows: if a module at such a row has a non-trivial source, or if a trivial source module appears at another row, the classification fails. A direct computational check on a small example, such as a block of a group with a cyclic Sylow $p$-subgroup of order $p^2$, would settle the claim.","tokens_in":22800,"feed_emoji":"🌳","tokens_out":10951,"duration_ms":176252,"temperature":0.7,"pith_summary":"The paper sets out to classify, completely and explicitly, all indecomposable trivial source modules (direct summands of permutation modules) in a block—an indecomposable summand of the modular group algebra—whose defect group is a non-trivial cyclic $p$-group. Such cyclic-defect blocks are among the most thoroughly understood blocks in modular representation theory, yet a full description of their trivial source modules had previously been missing. The main result is an exact placement in the stable Auslander-Reiten quiver: a module with vertex $D_i$ sits at distance $d_+(X)=\\ell_i p^{n-i}-1$ from the positive boundary, where $\\ell_i$ is an alternating sum computed from the endo-permutation module $W$ attached to the block's source algebra. The second main theorem then enumerates the modules themselves, listing the Brauer-tree path, direction, and multiplicity that characterise each trivial source module. If the classification is right, every block with a non-trivial cyclic defect group has a complete, explicit description of its $p$-permutation modules.","feed_headline":"All trivial source modules in cyclic defect blocks are classified","feed_subtitle":"The paper pins each such module to a path on the Brauer tree with a direction and a multiplicity.","key_machinery":"The load-bearing object is the pair attached to any cyclic-defect block: its strengthened Brauer tree (an ordinary Brauer tree with alternating signs on vertices) and an indecomposable capped endo-permutation $kD$-module $W$, which together determine the source algebra of $B$ up to equivalence. The proof reduces $B$ step by step to the simpler block $c$ of $C_G(D_1)$: Green correspondence preserves vertices, sources, and distances; induction from the inertia group preserves lengths, vertices, and sources; and the source algebra of $c$ is isomorphic to $\\operatorname{End}_k(W)\\otimes_k kD$. Under the resulting Morita equivalence, the unique trivial source $c$-module with vertex $D_i$ corresponds to the $kD$-module $U_{D_i}(W)=(\\operatorname{Ind}_{D_i}^{D}\\circ\\operatorname{Cap}\\circ\\operatorname{Res}_{D_i}^{D})(W)$, so its composition length is $\\ell_i p^{n-i}$. The paper computes this length by restricting the generators of the group of capped endo-permutation modules from $D$ to $D_i$ and summing the resulting alternating dimension series. Finally, Theorem 5.3 uses the path parametrisation of indecomposable modules together with published distance formulas for each path type, and selects exactly those whose distance to one of the two boundaries equals the value demanded by Theorem 5.1.","core_discovery":"The central discovery is two-part. Theorem 5.1 says that if $X$ is a non-projective indecomposable trivial source $B$-module with vertex $D_i$, then its distance to the positive boundary of the stable Auslander-Reiten quiver is $d_+(X)=\\ell_i p^{n-i}-1$, with $\\ell_i=\\sum_{0\\le i_j<i}(-1)^j p^{i-i_j}+(-1)^{|\\{j:0\\le i_j<i\\}|}$; here $\\ell_i$ is the dimension of the cap of the restriction of $W$ to $D_i$, and $W$ is the indecomposable capped endo-permutation module determined by the source algebra of $B$. Theorem 5.3 turns this numerical criterion into a full list: for $e>1$, a non-projective indecomposable module is a trivial source module with vertex $D_i$ exactly when it belongs to one of seven path types on the Brauer tree and its direction $\\varepsilon$ and multiplicity $\\mu$ satisfy one of two arithmetic conditions in terms of $\\ell_i p^{n-i}$ and $m$, the exceptional multiplicity. In the case $e=1$, each vertex $D_i$ carries exactly one trivial source module, a uniserial module whose length is $\\ell_i p^{n-i}$ or $p^n-\\ell_i p^{n-i}$ depending on the sign of the unique non-exceptional character. The classification is completed by a correction to the earlier list of liftable modules when the exceptional vertex is a leaf, and by the observation that cotrivial source modules are obtained from trivial source modules by the Heller operator.","pith_inferences":["Editorial inference: the classification yields an algorithm for listing all $p$-permutation modules of any cyclic-defect block directly from the signed Brauer tree and the coefficients $(a_1,\\ldots,a_{n-1})$ of $W$, without constructing the whole module category.","Editorial inference: since the row position depends only on the restriction of $W$ to $D_i$, the formula suggests a general dictionary between elements of the group of capped endo-permutation modules and rows of the stable Auslander-Reiten quiver; extending this dictionary to blocks with other defect groups would be a natural next step.","Editorial inference: the arithmetic conditions in Theorem 5.3 could be read as a character-theoretic criterion for a liftable module to be a summand of a permutation module, since trivial source modules are detected by non-negative integer character values on generators of $D_1$."],"forward_implications":["In every block with cyclic defect group $D=C_{p^n}$, the trivial source modules with vertex $D_i$ form an explicitly enumerated finite list; no search over the module category is needed.","For a principal block, every non-projective trivial source module with vertex $D_i$ lies at distance $p^{n-i}-1$ from the positive boundary (Corollary 5.2(d)).","The hooks of the block—the modules on the boundary of the stable Auslander-Reiten quiver—are trivial source modules exactly when the endo-permutation module $W$ is trivial and the hook's character is positive at a generator of $D_1$ (Corollary 5.2(c)).","The cotrivial source modules are classified by the same lists after replacing $\\ell_i p^{n-i}-1$ with $p^n-\\ell_i p^{n-i}$ (Remark 5.5).","The corrected enumeration of liftable modules (Appendix A) implies that in a block with $e=1$ every indecomposable module is liftable, and the unique trivial source module with vertex $D_i$ is uniserial of length either $\\ell_i p^{n-i}$ or $p^n-\\ell_i p^{n-i}$ according to the sign of the unique non-exceptional character (Theorem 5.3(a))."],"supporting_citations":[{"why":"Supplies the classification of indecomposable liftable modules that the new classification filters through, corrected in Appendix A for the exceptional-vertex-as-leaf case.","marker":"[HN12, Theorem 2.1]"},{"why":"Provides the path and top-socle parametrization of non-projective indecomposable modules used to label trivial source modules.","marker":"[Jan69, §5]"},{"why":"Determines distances from modules to the boundaries of the stable Auslander-Reiten quiver in terms of path, direction, and multiplicity, giving the formulas Theorem 5.3 tests against Theorem 5.1.","marker":"[BC02]"},{"why":"Shows the source algebra, hence the block up to source-algebra equivalence, is determined by the strengthened Brauer tree and the endo-permutation module $W$ used in the distance formula.","marker":"[Lin96, Theorem 2.7]"},{"why":"Gives the shape of the stable Auslander-Reiten quiver for cyclic blocks and the Green correspondence facts used to reduce distances from $B$ to its Brauer correspondent.","marker":"[Ben98, Theorem 6.5.5]"},{"why":"Classifies capped endo-permutation modules over cyclic $p$-groups, yielding the generators and dimension formulas needed to compute $\\ell_i$.","marker":"[Dad78]"},{"why":"Relates character values of the lift of a trivial source module to fixed points on $p$-subgroups, the criterion used to identify hook-type trivial source modules.","marker":"[Lan83, Lemma II.12.6]"},{"why":"Establishes that a single $\\Omega^2$-orbit in the stable quiver has vertex $D_i$ and trivial source, reducing the location problem to computing the row distance.","marker":"[Bes91]"}],"fun_headline_variants":["All trivial source modules in cyclic defect blocks now classified","Trivial source modules pinned to Brauer tree paths","Cyclic defect blocks: complete classification of trivial source modules","Trivial source modules in cyclic defect blocks fully classified","Brauer tree paths classify trivial source modules in cyclic defects"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The classification inherits its completeness from the corrected enumeration of all indecomposable liftable modules given in Appendix A: every trivial source module is liftable, so if that enumeration missed any liftable module—for example in the special situation where the exceptional vertex is a leaf of the Brauer tree—then the new classification of trivial source modules would be incomplete as well.","fun_headline_variants_meta":{"raw":{"variants":["All trivial source modules in cyclic defect blocks now classified","Trivial source modules pinned to Brauer tree paths","Cyclic defect blocks: complete classification of trivial source modules","Trivial source modules in cyclic defect blocks fully classified","Brauer tree paths classify trivial source modules in cyclic defects"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000184,"raw_usage":{"total_tokens":1326,"prompt_tokens":963,"completion_tokens":363,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":579,"completion_tokens_details":{"reasoning_tokens":285}},"tokens_in":579,"tokens_out":363,"duration_ms":593025,"temperature":1.0,"reasoning_tokens":285,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:55:30.978253+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"One concrete test: in a principal block with cyclic defect group $C_{p^n}$, the theorem predicts that every trivial source module with vertex $D_i$ lies exactly $p^{n-i}-1$ rows from the positive boundary of the stable Auslander-Reiten quiver. Compute that quiver explicitly and inspect the sources of the liftable modules at the predicted rows: if a module at such a row has a non-trivial source, or if a trivial source module appears at another row, the classification fails. A direct computational check on a small example, such as a block of a group with a cyclic Sylow $p$-subgroup of order $p^2$, would settle the claim.","supporting_citations":[],"review_version":1}