{"id":"c45f75cb-600a-40a5-9da7-8eeb97c736a9","arxiv_id":"2411.14814","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"The Albanese variety and Albanese fibers of a hyperelliptic variety X=A/G are explicitly described as (A0/K0)/G and ({a0}+A1)/H, yielding new indecomposability results for derived categories.","lead":"This paper gives a complete explicit description of the Albanese morphism for hyperelliptic varieties, spaces obtained by folding an abelian variety by a finite symmetry group. It also proves that the derived category is indecomposable for several large classes, including all hyperelliptic threefolds.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified: the proof of Theorem 26 is internally consistent and the construction of the G-stable complement A1 in Lemma 15 is secure.","rationale":"The reader correctly identified Lemma 15 as the most delicate assumption underlying the explicit description of the Albanese morphism. We focused our stress test on exactly that step and on the isogeny argument in Lemma 25/Theorem 26. Our analysis shows that the construction of A1 as the orthogonal complement with respect to a G-invariant polarization is sound: the rationality of V1 follows from its equality with the E-orthogonal complement of a rational subspace, and the G-stability follows from the invariance of the polarization class. The non-uniqueness issues are addressed by the careful proofs in Lemmas 19 and 23. The final step showing that the candidate morphism is the Albanese morphism is also valid: the dimension computation is correct and the connectedness of the fibers of f, established in Proposition 24, forces the isogeny to be an isomorphism. We found no internal inconsistency, no hidden circularity, and no unsupported leap in the central theorem. The applications, including the indecomposability results, rely on standard theorems and are consistent with the main construction. Therefore we agree with the reader's ACCEPT verdict and recommend no change. We mark agreement as 'partial' because, while we examined the same lemma the reader flagged, we do not regard it as an actual weakness: it is the most technical step, but it is rigorously proved.","tokens_in":22595,"tokens_out":23763,"duration_ms":223122,"concrete_test":"As a robustness check, independently recompute the decomposition A0, A1 and the resulting B=(A0/K0)/G for the non-cyclic example in Example 33 (G=(Z/3)^2) using a different G-invariant ample polarization from the one implicitly chosen in the paper. Verify that the quotient B is isomorphic to the Albanese variety Alb(X) and that the fibers are isomorphic to (E1×E2)/⟨g1⟩, confirming that the description in Theorem 26 is independent of the choice of complement up to isomorphism.","verdict_should_be":"UNCHANGED","load_bearing_attack":"We examined the central claim (Theorem 26) and its two key supporting steps: the decomposition A ≅ (A0×A1)/K with a G-stable abelian complement A1 (Lemma 15), and the proof that the morphism f of (33) is the Albanese morphism via the isogeny argument in Lemma 25 and Theorem 26. We find no gap. In Lemma 15, the subspace V1 defined as the H-orthogonal complement of V0 is rational because it equals the E-orthogonal complement of the rational subspace V0 with respect to the alternating form E representing c1(L). Hence Λ∩V1 is a lattice and A1 is an abelian subvariety. The G-invariance of H follows from the G-invariance of c1(L), since H(v,w)=E(iv,w)+iE(v,w) and the linear parts are C-linear. The non-uniqueness of the decomposition τ(g)=t_g^0+t_g^1 is handled correctly in Lemmas 19 and 23. The factorization in Lemma 25 is valid: dim Alb(X)=dim V0=dim B, and the connectedness of the fibers of f forces the isogeny to have connected fibers, hence to be an isomorphism. We therefore find no load-bearing objection to the central claim.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper gives an explicit description of the Albanese morphism of a hyperelliptic variety X = A/G. The central result, Theorem 26, identifies Alb(X) with (A0/K0)/G and describes each Albanese fiber as ({a0}+A1)/H, where (A0,A1) is a G-stable decomposition of A and H ⊂ G is an explicitly defined subgroup. The proof is carried out by elementary quotient computations and a careful treatment of the 1-cocycle defining the G-action. In Section 4 the authors study which abelian or hyperelliptic varieties can occur as Albanese fibers, including a no-go result for bielliptic surfaces in threefolds. Section 5 applies the description to prove indecomposability (and often stable indecomposability) of the derived category in several cases, giving partial evidence for Conjecture B.","tokens_in":24,"tokens_out":7837,"duration_ms":686431,"significance":"If correct, this is a substantial and useful structural result: it turns the Albanese morphism of a hyperelliptic variety into an effective construction from the defining data (A,G), and it makes the geography of Albanese fibers accessible. The proof of Theorem 26 is self-contained and explicit, with detailed lemmas (Lemmas 15, 19, 22, 23, 24) and a transparent quotient computation; no parameters are fitted and no conclusion is assumed. The derived-category applications are a genuine payoff, and the paper is honest about the limits of the method, e.g. in Remark 44 and Remark 49. The main external reliance is on standard classification results and tables in [17, 23], which are used for auxiliary geography statements rather than for the core theorem.","major_comments":[],"minor_comments":[{"comment":"The construction of A1 as the H-orthogonal complement of V0 would benefit from one explicit sentence justifying that V1 is rational with respect to Λ. Since Im(H) = E takes integral values on Λ and V0 is rational, V1 is the E-orthogonal complement of a rational subspace in a non-degenerate rational alternating form, hence rational; this is exactly what makes Λ1 = Λ ∩ V1 a lattice and A1 an abelian subvariety. The reader can fill this in, but the paper currently leaves it implicit.","section":"Lemma 15"},{"comment":"The step 'by Proposition 24, the fibers of f are connected, so the fibers of \\bar f must be connected as well' is terse. It would be clearer to write \\bar f^{-1}(b) = alb_X(f^{-1}(b)) because alb_X is surjective, so connectedness of f^{-1}(b) implies connectedness of \\bar f^{-1}(b) by continuity. This is the only compressed step in the proof of the main theorem.","section":"Lemma 25 and Theorem 26"},{"comment":"Several exclusions in the proof of Proposition 35 are justified only by 'inspecting the table' in [17, Tables 2 and 3] and by [7, Proposition 2.3]. For a no-go theorem that is meant to be verifiable, it would be helpful to reproduce the relevant rows of those tables or to give the explicit restrictions on A1 in the text. This is a reproducibility issue, not a correctness gap in the main theorem.","section":"Proposition 35"},{"comment":"The phrase 'which is moreover abelian' is confusing because a hyperelliptic variety cannot be an abelian variety (Lemma 12). The intended meaning is presumably that the holonomy group G = Z/2Z × Z/2Z is abelian; the sentence should be rephrased accordingly.","section":"Example 43"},{"comment":"The notation in Table 1 is mostly clear, but the meaning of K in the third column should be stated once in the caption, since earlier K is defined only in the surrounding text for the general construction.","section":"Table 1"}],"recommendation":"minor_revision","confidential_remarks":"I agree with the reader's positive assessment. The core theorem is sound and the proof is explicit; the classification-dependent parts of Section 4 rely on standard references and on the author's preprint [8], but those parts are auxiliary to the main claim. The paper is a good fit for the journal. The requested changes are local and presentational."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This is the paper that actually writes down the Albanese morphism for hyperelliptic varieties. Theorem 26 gives explicit formulas for Alb(X) and the fibers, upgrading Ueno's more opaque Theorem 7.13, and the applications to derived categories follow from the new description rather than from a new criterion. The main theorem is correct, and the threefold geography result (Proposition 35) is a genuine surprise.\n\nThe proof of Theorem 26 is the strongest part. The decomposition A ~ (A0 x A1)/K is handled carefully: Lemma 15 constructs the G-stable complement A1 via a G-invariant hermitian form, and the cocycle handling in Lemmas 19, 23, and 24 is consistent. The stress-test's second pass is right, no gap there. I also like that the paper is explicit about the non-uniqueness of the decomposition of the translation part and shows the final formulas are independent of choices.\n\nThe derived-category section is a correct application of known tools (Pirozhkov's stability criterion and Kawatani-Okawa). Corollaries 38, 40, and 47 are real; they cover all hyperelliptic threefolds, which is a nice state of affairs. Proposition 35 is the most interesting: not every bielliptic surface appears as an Albanese fiber of a hyperelliptic threefold. That is a genuine geographic statement.\n\nSoft spots: the no-go result leans on classification tables (Lange's [17], Catanese-Demleitner [5]). If those tables are solid, the argument is fine, but the paper doesn't reprove them. That is normal practice in this area, so I don't count it heavily. Proposition 46 (canonical bundle formula) is dense; I did not verify every line of the Fujino-Mori application, but the logic is standard. Minor: calling a point an abelian fiber might confuse some readers, but it is harmless.\n\nThe paper is written cleanly, definitions are precise, and the citation practice is honest: the authors explicitly credit Ueno's precursor and explain what is new. It is a good example of how a concrete structural result can unlock applications.\n\nBottom line: worth a serious referee. I would send it to a journal and let a specialist check the classification-dependent parts. Not desk-reject.","headline":"A solid, genuinely useful computation of the Albanese morphism for hyperelliptic varieties, with correct derived-category applications; deserves serious refereeing.","tokens_in":23377,"tokens_out":3115,"would_cite":true,"duration_ms":30796,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14K05","14F08"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper gives an explicit formula for the Albanese morphism of every hyperelliptic variety, describing both the Albanese variety and its fibers as subquotients of the covering abelian variety, and uses this to prove indecomposability…","keywords":["Albanese morphism","hyperelliptic varieties","abelian varieties","étale quotients","Albanese fibers","derived categories","semiorthogonal decompositions","irregularity"],"falsifier":"Compute the Albanese morphism for a concrete hyperelliptic threefold with holonomy group $D_4$ or the $\\mathbb{Z}/2\\mathbb{Z}\\times\\mathbb{Z}/2\\mathbb{Z}$ example in Example 43, and compare each fiber with the predicted quotient $({a_0}+A_1)/H$ from Theorem 26; finding any fiber not isomorphic to such a quotient, or an Albanese variety not isomorphic to $(A_0/K_0)/G$, would refute the main theorem.","tokens_in":2240,"feed_emoji":"📐","tokens_out":3190,"duration_ms":65507,"temperature":0.7,"pith_summary":"The paper establishes an explicit description of the Albanese morphism for hyperelliptic varieties, which are quotients of abelian varieties by finite groups acting freely and with a non-translation component. It proves that the Albanese variety is a quotient of a subquotient determined by the fixed part of the group action, and that every Albanese fiber is itself a quotient of a complementary abelian subvariety; hence each fiber is either an abelian variety or a hyperelliptic variety. This makes the Albanese fibration fully computable from the defining data $A$ and $G$. As an application, the derived category of coherent sheaves is indecomposable in several large classes, including all hyperelliptic threefolds, supporting the conjecture that this holds for all hyperelliptic varieties.","feed_headline":"Hyperelliptic Albanese fibers are explicit subquotients","feed_subtitle":"A formula using the covering abelian variety and its group action describes the Albanese map and proves derived-category indecomposability…","key_machinery":"The central mechanism is the isotypical decomposition of the abelian variety $A$ with respect to the linear representation of $G$: the subvariety $A_0$ of points fixed by all linear parts, a $G$-stable complement $A_1$ obtained as the orthogonal complement with respect to a $G$-invariant hermitian form, and the kernel $K$ of the addition isogeny $A_0\\times A_1\\to A$. The paper then tracks the translation part of each $g\\in G$ through this decomposition, producing the subgroup $H$ that acts on translates of $A_1$. The key identity is the fiber formula $f^{-1}([a_0+K_0]_G)=({a_0}+A_1)/H$, which converts the geometric Albanese fibration into a group-action subquotient computation.","core_discovery":"Given a hyperelliptic variety $X=A/G$ with $G$ acting freely and without translations, the paper constructs two abelian subvarieties $A_0,A_1\\subset A$: $A_0$ is where the linear parts of $G$ act trivially, and $A_1$ is a $G$-stable complement obtained from a $G$-invariant polarization, so that the addition map $A_0\\times A_1\\to A$ is an isogeny with kernel $K$. The paper proves that the Albanese morphism is the natural quotient map $X\\to (A_0/K_0)/G$, where $K_0$ is the projection of $K$ to $A_0$, and that for every $a_0\\in A_0$ the Albanese fiber is isomorphic to $({a_0}+A_1)/H$, with $H\\subseteq G$ the subgroup whose translation part on $A_0$ lies in $K_0$. The Albanese fibers are therefore abelian varieties or hyperelliptic varieties, and both the Albanese variety and the fibers are explicitly determined by $A$ and $G$. The proof shows that the candidate morphism has connected fibers and the same dimension as the Albanese variety, forcing the natural isogeny from the Albanese variety to the candidate to be an isomorphism.","pith_inferences":["Because the description is phrased entirely in terms of the complex representation of $G$ and the cocycle $\\tau$, it should be possible to compute Albanese data for any hyperelliptic variety from linear algebra alone, which may simplify higher-dimensional classifications.","The paper's Conjecture B remains open exactly at the cases where the Albanese fiber has nontrivial torsion canonical bundle and the irregularity is small, such as the regular hyperelliptic fourfolds with holonomy $D_4\\times\\mathbb{Z}/2\\mathbb{Z}$; this suggests the hardest cases are those where holonomy is purely non-translational.","Proposition 35 indicates that the set of lower-dimensional hyperelliptic varieties appearing as Albanese fibers is a proper and group-theoretically constrained class, so the geography of Albanese fibers is a tractable classification problem in its own right.","The canonical-bundle argument in Proposition 46 may extend to show $[\\omega_X]\\in\\operatorname{Pic}^0(X)$ whenever the Albanese morphism has any fiber with trivial canonical bundle, not only when all fibers are Calabi-Yau; this would be a natural strengthening of the indecomposability criterion."],"forward_implications":["For any hyperelliptic variety $X=A/G$, both the Albanese variety and every Albanese fiber can be computed from the representation of $G$ on $A$ and the cocycle $\\tau$, making the Albanese fibration explicit and algorithmic.","The Albanese fibers are severely constrained: each is either an abelian variety or a lower-dimensional hyperelliptic variety, never an arbitrary variety of Kodaira dimension zero.","When $X$ is cyclic, the fibers are abelian or cyclic hyperelliptic varieties, and under an eigenvalue condition on the generator the fibers are literally the abelian variety $A_1$.","The explicit description yields a no-go result for hyperelliptic threefolds: a general member of the $1$-dimensional family of bielliptic surfaces with holonomy $\\mathbb{Z}/3\\mathbb{Z}$ does not occur as an Albanese fiber.","The derived category $\\mathrm{D}^b(X)$ is indecomposable for cyclic hyperelliptic varieties, for varieties of irregularity $\\dim X-1$ or $\\dim X-2$, and for varieties whose Albanese fibers have trivial canonical bundle; in particular all hyperelliptic threefolds have indecomposable derived categories."],"supporting_citations":[{"why":"Supplies the theorem that the Albanese morphism of a variety with torsion canonical bundle is an étale fiber bundle with smooth connected fibers, which underlies Lemma 14.","marker":"[12]"},{"why":"Provides the earlier quotient description of Albanese morphisms for quotients of abelian varieties by arbitrary group actions, which the paper makes explicit in the hyperelliptic case.","marker":"[24]"},{"why":"Gives Poincaré complete reducibility and the hermitian-form/polarization formalism used to construct the $G$-stable complement $A_1$ in Lemma 15.","marker":"[3]"},{"why":"Introduces hyperelliptic varieties, the cocycle $\\tau$, and the classification tables for cyclic and low-dimensional cases used in Sections 3 and 4.","marker":"[17]"},{"why":"Provides the stable indecomposability criterion (Pirozhkov's theorem) used to prove Corollaries 38 and 40.","marker":"[22]"},{"why":"Supplies the Kawatani-Okawa indecomposability criterion for varieties with $[\\omega_X]\\in\\operatorname{Pic}^0(X)$, used in Corollary 47.","marker":"[13]"},{"why":"Lists the possible groups acting on $\\mathbb{C}^3$ for hyperelliptic threefolds, which is essential for the exclusion argument in Proposition 35.","marker":"[23]"},{"why":"Classifies hyperelliptic threefolds and provides examples used in Proposition 35 and in the regular threefold examples.","marker":"[5]"},{"why":"Classifies Bagnera-de Franchis varieties in small dimensions, including fixed-point assertions used in Example 33 and Proposition 35.","marker":"[7]"}],"fun_headline_variants":["Albanese morphism for hyperelliptic varieties made explicit","Hyperelliptic Albanese fibers: exact subquotients from group action","Explicit Albanese map: abelian and hyperelliptic fibers","Describing Albanese fibers of hyperelliptic varieties","Hyperelliptic Albanese: fibers are explicit subquotients"],"cache_read_input_tokens":25472,"weakest_assumption_plain":"The proof requires that the abelian variety $A$ admit a $G$-stable complementary subvariety $A_1$ to $A_0$, obtained as the orthogonal complement with respect to a $G$-invariant polarization; if no such complement exists, the formulas for the Albanese variety and its fibers would not be well-defined.","fun_headline_variants_meta":{"raw":{"variants":["Albanese morphism for hyperelliptic varieties made explicit","Hyperelliptic Albanese fibers: exact subquotients from group action","Explicit Albanese map: abelian and hyperelliptic fibers","Describing Albanese fibers of hyperelliptic varieties","Hyperelliptic Albanese: fibers are explicit subquotients"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000638,"raw_usage":{"total_tokens":2925,"prompt_tokens":915,"completion_tokens":2010,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":531,"completion_tokens_details":{"reasoning_tokens":1937}},"tokens_in":531,"tokens_out":2010,"duration_ms":15043,"temperature":1.0,"reasoning_tokens":1937,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T14:52:33.526286+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the Albanese morphism for a concrete hyperelliptic threefold with holonomy group $D_4$ or the $\\mathbb{Z}/2\\mathbb{Z}\\times\\mathbb{Z}/2\\mathbb{Z}$ example in Example 43, and compare each fiber with the predicted quotient $({a_0}+A_1)/H$ from Theorem 26; finding any fiber not isomorphic to such a quotient, or an Albanese variety not isomorphic to $(A_0/K_0)/G$, would refute the main theorem.","supporting_citations":[{"cited_title":"Classification of algebraic varieties. I","cited_arxiv_id":null,"evidence_quote":"Provides the earlier quotient description of Albanese morphisms for quotients of abelian varieties by arbitrary group actions, which the paper makes explicit in the hyperelliptic case."},{"cited_title":"Stably semiorthogonally indecomposable varieties","cited_arxiv_id":null,"evidence_quote":"Provides the stable indecomposability criterion (Pirozhkov's theorem) used to prove Corollaries 38 and 40."},{"cited_title":"Nonexistence of semiorthogonal decompositions and sections of the canonical bundle","cited_arxiv_id":"1508.00682","evidence_quote":"Supplies the Kawatani-Okawa indecomposability criterion for varieties with $[\\omega_X]\\in\\operatorname{Pic}^0(X)$, used in Corollary 47."},{"cited_title":"The classification of hyperelliptic threefolds","cited_arxiv_id":null,"evidence_quote":"Classifies hyperelliptic threefolds and provides examples used in Proposition 35 and in the regular threefold examples."},{"cited_title":"Classification of Bagnera–de Franchis varieties in small dimensions","cited_arxiv_id":null,"evidence_quote":"Classifies Bagnera-de Franchis varieties in small dimensions, including fixed-point assertions used in Example 33 and Proposition 35."}],"review_version":1}