{"id":"f63c62cc-924e-4bda-903a-6557b2f2bc88","arxiv_id":"2501.06629","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"In a finite tensor category, an algebra is exact if and only if it has no nonzero nilpotent ideals, equivalently if and only if it is a finite product of simple algebras.","lead":"Mathematicians proved that a certain class of algebras, called exact algebras, inside finite tensor categories are exactly the finite products of simple building blocks, confirming a conjecture by Etingof and Ostrik. The proof introduces a new radical construction that encodes when an algebra is exact, with applications to tensor categories coming from vertex operator algebras and symmetric categories.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The proof of Theorem 7.1 rests on Proposition 6.13, which imports [St2, Lemma 4.16] and [St2, Lemma 4.11]; the positivity argument for the Perron-Frobenius idempotent in Remark 6.14 is the point most worth independent verification.","rationale":"I read the paper in good faith and found no definite error in the central argument. The main theorem is a chain of correspondences: ideals of A ↔ mixed subfunctors ↔ C_p-stable ideals in Kl C_p⊗A (Theorem 5.9), followed by the radical construction and the exactness criterion. The most delicate step is Proposition 6.13, exactly as the Reader identified. It is the only place where an external result [St2] is used to show that the first map of a minimal projective presentation lands in the C-stable radical, and the authors' Remark 6.14 addresses the transitivity hypothesis by an argument about the Perron-Frobenius idempotent. I agree that this is the most load-bearing imported input. I also noticed a smaller omission in Proposition 6.12: the proof shows that every nilpotent C_p-stable ideal is contained in Rad_C_A but does not explicitly prove Rad_C_A itself is nilpotent; however Lemma 6.11 (Rad_C_A ⊆ Rad Kl C_p⊗A) combined with nilpotency of the ordinary radical of the Hom-finite category Kl C_p⊗A supplies the missing line, so this is not a serious defect. The diagrammatic proofs in Sections 5 and 6 are dense but internally coherent; I did not find a concrete contradiction in them. Because the identified concern is precisely the Reader's weakest assumption, and because the authors have already stated the required positivity condition explicitly, my read does not change the verdict. The paper should still be accepted, ideally with an independent verification of the [St2] lemmas and the positivity claim in Remark 6.14.","tokens_in":25028,"tokens_out":13836,"duration_ms":138967,"concrete_test":"Independently re-derive Proposition 6.13 without importing [St2] verbatim: for a finite tensor category C and a nonzero R∈proj_C(A), prove that the Perron-Frobenius idempotent e∈R⊗Z[C_p]⊕ has strictly positive coefficient on every indecomposable projective object and that e⊳v≠0 for every nonzero non-negative v∈R⊗Z[proj_C(A)]⊕, then repeat the minimal-presentation argument. As a computational spot-check, instantiate C as the principal block of representations of the small quantum group at a root of unity (or another non-semisimple finite tensor category), compute the matrix of the action of e on all indecomposable projective A-modules, and verify that no nonzero non-negative vector is annihilated; if such a vector exists, Proposition 6.13 is false.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem 6.16 reduces exactness of A to vanishing of Rad_C(A), and the nontrivial direction (iii)⇒(i) passes through Proposition 6.13: for a minimal projective presentation R1→R0→Q⊳X, the morphism r1 lies in the C-stable radical (Rad_C_A)^c. This is the only step where absence of nilpotent ideals is converted into a statement about all projective modules over A, so if Proposition 6.13 fails the equivalence between exactness and radical vanishing collapses. The proof invokes [St2, Lemma 4.16] and [St2, Lemma 4.11], and the authors argue in Remark 6.14 that the transitivity condition demanded there is automatically satisfied. Their argument reduces the condition to the claim that the Perron-Frobenius idempotent e in R⊗Z[C_p]⊕ is a positive linear combination of all indecomposable projectives, and that {P⊳R | P∈C_p} ≠ {0} for every nonzero R∈proj_C(A). This is plausible and probably true, but it is the least independently supported part of the proof: the cited lemmas are not reproduced, the positivity of e is not proved in the text, and no formalization is provided. A failure of positivity, or a nonzero projective A-module annihilated by all P∈C_p, would make r1 non-radical and would break the implication from no nilpotent ideals to exactness. The rest of the main theorem is comparatively elementary, so this is the single load-bearing spot.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper proves Etingof and Ostrik's Conjecture B.6: an algebra object A in a finite tensor category C is exact if and only if it is a finite direct product of simple algebras. The proof introduces mixed subfunctors of C(-,A) restricted to projective objects, establishes a lattice isomorphism between ideals of A and C_p-stable ideals in the Kleisli category Kl(C_p⊗A) via two independent routes (a pseudo-Hopf algebra argument in Section 3 and a Day-convolution argument in Section 4), and uses this to define a C-module radical Rad_C(A). The main theorem characterizes exactness by Rad_C(A)=0, by absence of nonzero nilpotent ideals, and by decomposability into simple algebras. The paper also derives consequences for the module category of A/Rad_C(A) and for incompressible finite symmetric tensor categories.","tokens_in":25286,"tokens_out":16111,"duration_ms":128944,"significance":"If correct, the paper resolves a conjecture of Etingof and Ostrik (and the equivalent conjecture of Shimizu) that had been open for finite tensor categories beyond the Hopf-algebraic setting. The main theorem is clean and the proof is well structured. Notably, the crucial bijection between algebra ideals and C_p-stable ideals is supported by two independent derivations, which cross-check each other. The paper also contains a self-contained proof of the known direction that exact algebras decompose into simples (Proposition 7.4), a new Jacobson-radical analogue with applications (Theorem 8.9), and concrete applications to incompressible symmetric tensor categories (Appendix B). These features make the paper a substantial contribution to the theory of module categories over finite tensor categories.","major_comments":[{"comment":"The proof of Proposition 6.12 shows only that every nilpotent C_p-stable ideal is contained in Rad^C_A; it does not show that Rad^C_A is itself nilpotent. The claim that Rad^C_A is the greatest nilpotent C_p-stable ideal, and consequently that Rad_C(A) is nilpotent, is needed for the equivalence (ii)⇔(iii) in Theorem 7.1 and for Lemma 8.8. The missing argument is short: because Kl(C_p⊗A) is Hom-finite with finitely many indecomposables, its Jacobson radical is nilpotent, and Lemma 6.11 gives Rad^C_A ⊆ Rad(Kl(C_p⊗A)), hence Rad^C_A is nilpotent. Please add this justification.","section":"§6.2 (Proposition 6.12)"},{"comment":"The proof of Proposition 6.13 imports [St2, Lemma 4.16] and [St2, Lemma 4.11] and argues that the transitivity condition in loc. cit. reduces to the positivity of the Perron–Frobenius idempotent e and to {P⊳R | P∈C_p} ≠ {0}. The non-vanishing condition is indeed immediate, but the positivity of e is only asserted ('by construction'), and the precise way in which [St2, Lemma 4.16] yields a positive idempotent without the transitivity assumption is not demonstrated. Since this step converts the absence of nilpotent ideals into the vanishing of the projective presentation morphism r_1, it is load-bearing. Please spell out the relevant statement from [St2] or provide a direct proof.","section":"§6.2 (Proposition 6.13, Remark 6.14)"}],"minor_comments":[{"comment":"In Remark 6.14, consider adding a one-sentence justification of the non-vanishing condition: if R is a nonzero projective A-module, then for a projective cover Q ↠ 1 in C, the map Q⊳R ↠ R is a nonzero epimorphism, so {P⊳R | P∈C_p} ≠ {0}.","section":"§6.2 (Remark 6.14)"},{"comment":"The section title 'The radical of a module object and the maximal semisimple qotient' contains a typo: 'qotient' should be 'quotient'.","section":"§8 (section title)"},{"comment":"In the proof of Proposition 6.13, the variables S and M are introduced 'using the notation of [St2]' but are not used further in the proof; aligning the notation with the statement of [St2, Lemma 4.16] would improve readability.","section":"§6.2 (proof of Proposition 6.13)"}],"recommendation":"major_revision","confidential_remarks":"The paper is a strong resolution of a known conjecture, and the two independent approaches to the key bijection inspire confidence. The main gap I identified (Proposition 6.12) is local and easily fixable. The reliance on [St2] is heavy, but the authors are aware of the delicate point; my second major comment asks them to make that step more explicit. I recommend major revision rather than rejection because the central claim is sound and the issues are local."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The conjecture is settled, and the proof is as good as the claim. The main theorem — exact iff finite product of simple algebras — is a real advance: before this, the converse was known only under a fibre functor, via Skryabin. The paper gives two independent routes to the key bijection between ideals in A, mixed subfunctors, and C_p-stable ideals (pseudo-Hopf in Section 3, Day convolution in Section 4), then builds a Jacobson-style radical that makes exactness equivalent to vanishing of the radical and to absence of nilpotent ideals. That chain is coherent and genuinely proves the conjecture, not just a reformulation. The applications to incompressible symmetric tensor categories in Appendix B are a nice payoff, and the note connecting the result to Shimizu's quasi-Frobenius conjecture is correct. I also appreciate that the authors are explicit about where they import heavy machinery.\n\nThe soft spot is exactly where the reader's stress-test points: Proposition 6.13, the step that converts 'no nilpotent ideals' into a statement about all projective A-modules, rests on two lemmas from Stroinski's paper plus a Perron–Frobenius idempotent whose positivity is asserted rather than proved in the text. Remark 6.14 gives a plausible reduction: finitely many indecomposable projectives, positivity of the idempotent, and the condition that {P ⊳ R} is non-zero for every non-zero projective R. That is probably true, but it is not self-contained, and the cited lemmas are not reproduced. If a specialist found a counterexample to the positivity claim, the implication from (iii) to (i) in Theorem 7.1 would collapse. I would not call this a fatal gap — the argument looks right and the authors clearly thought about the transitivity condition — but it is the load-bearing imported input, and the referee report should ask for a more detailed proof of the positivity claim or at least a precise statement of how [St2, Lemma 4.16] applies.\n\nThe rest of the main theorem is comparatively elementary once the bijection and radical are in place, and the diagrammatic proofs are dense but checkable. No circularity: the radical is defined via the bijection, but the equivalence with absence of nilpotent ideals is proved. No fitted parameters, no invented target.\n\nNet: this deserves a serious referee. I would send it out, and if the Proposition 6.13 point survives scrutiny, accept. For my own reading group, I would bring it up. I would also cite it in the next twelve months.\n\nRecommendation: peer review, with focused attention on Section 6.2 and the imported [St2] lemmas.","headline":"Settles Etingof–Ostrik's Conjecture B.6 with a coherent, well-supported proof; the one point a referee should press is the imported Perron–Frobenius argument in Proposition 6.13.","tokens_in":25880,"tokens_out":1514,"would_cite":true,"duration_ms":17052,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["18D25","18M05"],"pacs":[],"model":"deepseek-v4-flash","headline":"Exact algebras are exactly finite products of simple ones.","keywords":["exact algebra","finite tensor category","simple algebra","Jacobson radical","module category","C-module radical","mixed subfunctor","nilpotent ideal"],"falsifier":"Compute whether the module category of $A \\otimes A$ over $\\mathbb{Z}/2$-graded vector spaces in characteristic 2 contains a non-projective object, where $A = k[x]/(x^2)$ with $x$ odd; the theorem predicts it must, since $A \\otimes A$ is not a finite product of simple algebras, so finding every module projective would falsify the converse direction.","tokens_in":24756,"feed_emoji":"🧮","tokens_out":8551,"duration_ms":73128,"temperature":0.7,"pith_summary":"This paper claims to settle a conjecture from the literature on finite tensor categories: an algebra object is exact if and only if it is a finite direct product of simple algebras. Exactness here means that every module over the algebra is projective in the categorical sense, a property known to hold for products of simples but open in general. The proof introduces a categorical analogue of the Jacobson radical for algebra objects, called the C-module radical, and shows that an algebra is exact precisely when this radical vanishes. This gives a concrete criterion: an algebra has no nonzero nilpotent ideals exactly when it is a finite product of simple algebras. A sympathetic reader should care because this unifies several prior results and settles the structural question for module categories over finite tensor categories.","feed_headline":"Exact algebras are exactly finite products of simple ones","feed_subtitle":"A new radical for algebra objects settles a long-open structural conjecture.","key_machinery":"The load-bearing device is a chain of bijections between three kinds of ideals: two-sided ideals inside the algebra object $A$, mixed subfunctors of the representable presheaf $\\mathcal{C}(-,A)$ restricted to projective objects, and $\\mathcal{C}_p$-stable ideals in the Kleisli category $\\mathrm{Kl}(\\mathcal{C}_p \\otimes A)$ of free projective $A$-modules. The mixed subfunctor notion is new here; it packages the data of a subfunctor that is simultaneously compatible with both left and right multiplication. The C-module radical $\\mathrm{Rad}_{\\mathcal{C}}(A)$ is defined as the ideal corresponding to the largest nilpotent $\\mathcal{C}_p$-stable ideal in the Kleisli category, and its vanishing is shown to be equivalent to exactness using minimal projective presentations and a Perron–Frobenius type idempotent imported from earlier work.","core_discovery":"The central result is Theorem 7.1: for an algebra object $A$ in a finite tensor category $\\mathcal{C}$, four conditions coincide — $A$ is exact; the new radical $\\mathrm{Rad}_{\\mathcal{C}}(A)$ is zero; $A$ has no nonzero nilpotent ideal objects; and $A$ is a finite direct product of simple algebras. The paper proves the missing direction: any finite product of simple algebras is exact, so exact algebras are precisely the semisimple-like objects. Along the way it establishes a canonical bijection between ideals in $A$ and certain stable ideals in the Kleisli category of free projective $A$-modules, and it shows that $\\mathrm{Rad}_{\\mathcal{C}}(A)$ is the largest nilpotent ideal. The result also verifies an equivalent conjecture about quasi-Frobenius algebras in finite tensor categories.","pith_inferences":["If the radical theory behaves like the classical one, $\\mathrm{Rad}_{\\mathcal{C}}(A)$ could be used to measure how far $\\mathrm{mod}_{\\mathcal{C}}(A)$ is from being semisimple, and the quotient $A/\\mathrm{Rad}_{\\mathcal{C}}(A)$ might play the role of the maximal semisimple quotient in tensor-categorical module theory.","The bijection between ideals in $A$ and $\\mathcal{C}_p$-stable ideals suggests that lattice-theoretic questions about module categories reduce to ideal theory in the algebra object; one could test this by computing ideal lattices in examples such as the symmetric category of $\\mathbb{Z}/2$-graded vector spaces in characteristic 2.","Remark 6.15 of the paper indicates the radical method might extend to non-finite tensor categories with enough projectives, such as those from Lie superalgebras, except where the finiteness-based Proposition 6.13 fails; finding a concrete category where the Perron–Frobenius idempotent exists but finiteness fails would probe that boundary."],"forward_implications":["Every simple commutative algebra in a braided finite tensor category gives rise to a finite tensor category of modules, since it is exact.","In finite symmetric tensor categories, there is a unique largest simple commutative algebra $F(\\mathcal{C})$, and the category of its modules is the unique incompressible quotient of $\\mathcal{C}$.","The category of modules over $A/\\mathrm{Rad}_{\\mathcal{C}}(A)$ consists exactly of subquotients of objects $Q \\rhd L$ with $Q$ projective and $L$ semisimple, giving a maximal-semisimple-quotient picture.","Without assuming a fibre functor, every module over a simple algebra is projective in the categorical sense, extending a classical Hopf-algebra result to arbitrary finite tensor categories.","The equivalence of exactness with having no nonzero nilpotent ideals gives a practical test for exactness of algebra objects."],"supporting_citations":[{"why":"Formulates the conjecture (Conjecture B.6) that this paper proves, and supplies the pseudo-Hopf algebra realisation used in Section 3.","marker":"[EO2]"},{"why":"Provides the Perron–Frobenius type idempotent and the radical-inclusion lemmas (Lemma 4.16 and Lemma 4.11) that carry the key implication from exactness to vanishing of the radical.","marker":"[St2]"},{"why":"The classical Hopf module algebra theorem whose tensor-categorical formulation the paper extends beyond the fibre-functor setting.","marker":"[Sk]"},{"why":"States the quasi-Frobenius conjecture for simple algebras, shown in the paper to be equivalent to the main conjecture and therefore proven here.","marker":"[Sh]"},{"why":"Its Lemma 12 connects stable ideals with exactness of module categories; the paper generalises this connection via minimal projective presentations.","marker":"[MM]"},{"why":"Supplies the idempotent-closure and Tambara-module facts used to transfer between ideals in the Kleisli category and projective module categories.","marker":"[St1]"}],"fun_headline_variants":["Exact algebras are exactly finite products of simple ones","New radical for algebras proves Etingof–Ostrik conjecture","Algebra objects: exact iff finite direct product of simples","Jacobson-radical analogue settles exact algebra structure"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a Perron–Frobenius type idempotent exists in the split Grothendieck semiring of projective objects, so that the imported lemmas apply; if that fails, the proof that exactness forces the new radical to vanish would break.","fun_headline_variants_meta":{"raw":{"variants":["Exact algebras are exactly finite products of simple ones","New radical for algebras proves Etingof–Ostrik conjecture","Algebra objects: exact iff finite direct product of simples","Jacobson-radical analogue settles exact algebra structure"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000153,"raw_usage":{"total_tokens":1112,"prompt_tokens":758,"completion_tokens":354,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":374,"completion_tokens_details":{"reasoning_tokens":289}},"tokens_in":374,"tokens_out":354,"duration_ms":4024,"temperature":1.0,"reasoning_tokens":289,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T20:56:54.363692+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute whether the module category of $A \\otimes A$ over $\\mathbb{Z}/2$-graded vector spaces in characteristic 2 contains a non-projective object, where $A = k[x]/(x^2)$ with $x$ odd; the theorem predicts it must, since $A \\otimes A$ is not a finite product of simple algebras, so finding every module projective would falsify the converse direction.","supporting_citations":[],"review_version":1}