{"id":"0ac51e11-7360-4b79-9cfb-6a13c922d2c3","arxiv_id":"2601.21871","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Polync varieties generalize normal-crossing degenerations; the paper proves d-semistable K-trivial polync varieties are smoothable and constructs multiparameter Kulikov models with explicit examples.","lead":"Using a new class of \"polync\" singular spaces, the paper proves that certain singular degenerations with trivial canonical bundle can be smoothed, and builds explicit multiparameter families of K3 degenerations. Generalists should read it as the latest step in turning logarithmic deformation theory into concrete combinatorial tools for studying degenerations of Calabi-Yau spaces.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 2.5 rests on the unproven forward direction of Prop. 2.2 (d-semistability ⇒ existence of a saturated log structure of semistable type); a gap in this multi-color adaptation would block the log-BTT input.","rationale":"I read the paper's central claim as Theorem 2.5, a multiparameter smoothability theorem derived by applying the logarithmic Bogomolov–Tian–Todorov theorem. The reader identified Proposition 2.2 as the weakest assumption, and I agree that this is where the argument is least secure. However, I would sharpen the concern: the theorem only needs the forward direction of Prop. 2.2 (d-semistability implies existence of a log structure of semistable type), not the converse. The reverse direction is used in the interpretative Remark 2.6 and in the claim of equivalence, but it is not load-bearing for the proof of Theorem 2.5. The forward direction is sketched with references to Kato's local transition-system construction, and the multi-color generalization introduces potential global cocycle issues on strata of mixed color; these are asserted but not fully demonstrated. Proposition 2.3 (log smoothness and saturation) is also a one-line citation, and the non-standard base log structure 0_S makes it worth checking. I do not see an internal inconsistency in the lemmas presented, and the examples are plausible, so a CONDITIONAL verdict with medium risk is appropriate. The proposed concrete test—explicit verification on the simplest two-color example—would either expose a hidden obstruction or lend support to the forward direction. No change to the reader's verdict is needed.","tokens_in":17314,"tokens_out":26507,"duration_ms":301412,"concrete_test":"Independently re-derive the forward direction of Prop. 2.2 for the product of two nodal curves X0=C1×C2 (the simplest |S|=2 polync variety, dual complex a square): compute the sheaves L_red,L_blue explicitly, follow Kato's transition-system construction on the four double curves and the triple points, and check that the cocycle condition and the product-of-units condition can be satisfied simultaneously for both colors. Then verify that the glued log structure is saturated and log smooth over 0_S. If this case fails, Theorem 2.5 needs an additional hypothesis; if it succeeds, the adaptation is plausible but should still be written out in full.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is Theorem 2.5, which applies the log-BTT theorem to a log structure of semistable type on X0. That log structure is supplied by Proposition 2.2. The reader flags the omitted reverse direction, but for Theorem 2.5 the load-bearing part is the forward direction: d-semistability ⇒ existence of a log structure of semistable type. That proof is only a sketch, said to follow Kato 'nearly verbatim'. The critical unverified steps are: (i) the simultaneous modification of transition systems for all colors on overlaps of strata of mixed color, so that the product-of-units condition Q_j u_s^{(j),λμ}=1 holds globally; (ii) the assertion in Prop. 2.3 that the resulting log structure is log smooth and saturated over the non-standard log point 0_S with ghost N^S. If either step fails, the log-BTT theorem cannot be applied, and neither the unobstructedness nor the semistable smoothing conclusion follows. The omitted reverse direction is less central; it would affect the 'if and only if' statement and Remark 2.6, but not the existence direction of the smoothing theorem. Nevertheless, the forward direction itself is not fully written out or machine-checked, so the central theorem inherits risk from it.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces polync varieties (étale/analytic products of normal-crossing singularities) with an S-coloring of their dual polysimplicial complex, and defines d-semistability for such varieties. The central theorem (Theorem 2.5) asserts that a proper, d-semistable, S-colored, polync variety with trivial dualizing sheaf has unobstructed log smooth deformations over the log point 0_S, yielding formal and analytic semistable smoothings over a multiparameter base, as a generalization of Friedman's and Kawamata–Namikawa's smoothing theorems. The proof is via the logarithmic Bogomolov–Tian–Todorov theorem, using a log structure of semistable type supplied by Proposition 2.2. Sections 3–4 specialize to K-trivial surfaces: dual complexes, numerical d-semistability, period maps, parameter counts, monodromy cones, and explicit multiparameter Kulikov models, including a rhombicuboctahedron example and degenerations of elliptic K3 surfaces.","tokens_in":17642,"tokens_out":8650,"duration_ms":92095,"significance":"If the main smoothing theorem and the supporting log-structure dictionary are fully established, this would be a substantial extension of classical 1-parameter smoothing theory to higher-dimensional bases, with potential applications to moduli and degeneration theory. The surface-theoretic framework (slabs, monodromy cones, conservation of charge) is attractive and the examples are informative. However, the current manuscript leaves several load-bearing statements unproved or explicitly deferred—most notably the forward direction of Proposition 2.2, the log smoothness/saturation claim in Proposition 2.3, Proposition 3.10, and Proposition 3.15(5)—so the significance is conditional on these gaps being filled.","major_comments":[{"comment":"The forward direction (d-semistability ⇒ existence of a log structure of semistable type) is the input to the log BTT theorem and therefore to Theorem 2.5. The proof is only a sketch: the simultaneous modification of transition systems on overlaps of mixed-color strata, so that the product-of-units condition ∏_j u^{(j),λμ}_s = 1 holds globally for every color s, is asserted without detailed verification. The statement that the log structures of different colors 'do not interact at all' needs justification on overlaps where multiple colors appear, and the cocycle condition on triple overlaps is not proved for the multi-color case. A complete proof, or a precise reduction to [23] with all multi-color modifications spelled out, is required before Theorem 2.5 can be regarded as fully established.","section":"§2, Proposition 2.2"},{"comment":"It is claimed that the log structure of semistable type provided by Proposition 2.2 is log smooth and saturated, with proof reduced to [25, Thm. 3.5] and the definition of saturated morphisms. This is a critical step because the log BTT theorem of [15] requires a saturated log smooth morphism over the log point. The hypotheses of [25, Thm. 3.5] are not verified explicitly; in particular, it is not immediate that the local charts on a polync variety with mixed colors define a saturated log structure. Please expand the proof so the reader can see precisely how the cited theorem applies.","section":"§2, Proposition 2.3"},{"comment":"Proposition 3.10 characterizes d-semistability of polysnc Type III K3 surfaces in terms of the period φ_{X0} vanishing on slab classes. The proof is omitted ('similar to the snc case'), yet the proposition is used in Corollary 3.12 to assert the existence of a d-semistable locally trivial deformation, which in turn justifies why the examples in Section 4 are smoothable. Since this is a central step in the surface theory, the authors should provide a proof or a precise reference that covers the polync setting.","section":"§3, Proposition 3.10 and Corollary 3.12"},{"comment":"Part (5) of Proposition 3.15 is explicitly deferred to future work, but Example 4.2 uses it to assert that the universal d-semistable deformation of the rhombicuboctahedron surface is X → Δ^4 × Δ^16, with Δ^16 the locally trivial d-semistable deformations. The dimension count in Proposition 3.14 is also 'naive' and is explicitly conditional on the surjectivity statement that (5) supplies. As written, the dimension claims in Example 4.2 are therefore unsupported. The authors should either prove (5), weaken the example to a conditional statement, or find an independent argument for the dimension.","section":"§3, Proposition 3.15(5) and Example 4.2"}],"minor_comments":[{"comment":"The definition of 'polysimplicial complex' uses 'polyhedral complex' in the same sentence; consistent terminology would help.","section":"§1, Definition 1.6"},{"comment":"The notation T^0 for the tangent space of Spec C[[u_s]] is unconventional; consider using Der(C[[u_s]], C) or the usual tangent space.","section":"§1, Proposition 1.16"},{"comment":"Exercise 3.5 is stated as an exercise but is used later (e.g., in Corollary 3.12). It should be labeled as a lemma or its proof sketched.","section":"§3, Exercise 3.5"},{"comment":"The definition of Q(V,D) = χ_top(V\\D) is clear, but the term 'charge' is not standard; a brief comment on its origin would help.","section":"§3, Proposition 3.13"},{"comment":"The claim that over V(u_red) the fiber is an snc Type III K3 surface 'in the same locally trivial deformation class as the octahedral surface' would benefit from a brief justification or a reference to the relevant figure.","section":"§4, Example 4.2"},{"comment":"The paper uses 'polysnc' and 'polync' interchangeably in places; standardize the terminology after first definition.","section":"General"},{"comment":"The sentence 'The argument of Proposition 3.3 is quite general' is a comment on the proof and would read better in the proof itself.","section":"§3, Proposition 3.3"}],"recommendation":"major_revision","confidential_remarks":"The paper is ambitious and the results, if fully proved, would be a significant contribution. My recommendation is based on the load-bearing gaps in Proposition 2.2 (forward direction), Proposition 2.3, Proposition 3.10, and Proposition 3.15(5). These are fixable in principle, but as written they prevent the main theorem and the examples from being fully rigorous. I would be willing to revisit a revised version that provides complete proofs for these statements or clearly marks the affected results as conditional."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nHere's my take on Engel's polync paper. The main thing you should know: it introduces a genuinely new combinatorial framework for multiparameter degenerations, with concrete examples that look right, but the central smoothing theorem leans on a log-structure existence step that is only sketched, and two supporting results are deferred. That combination explains the conditional verdict.\n\nWhat is actually new: polync varieties, S-colorings, d-semistability, and the multiparameter Kulikov model formalism. These definitions are natural and the examples in Section 4 are the real payoff: the rhombicuboctahedron gives a 4-parameter Kulikov model with a computed monodromy cone, and the 10-color construction gives an explicit degeneration of elliptic K3 surfaces with a non-symplectic involution. The paper also credits prior work properly and states its own gaps.\n\nThe soft spot is the forward direction of Proposition 2.2. Theorem 2.5 applies the log BTT theorem to a log structure of semistable type, and that log structure is supposed to come from d-semistability. The proof is a sketch adapted from Kato, with the multi-color transition systems only outlined. If that step fails, the log-BTT input collapses. The reverse direction is omitted, but that is less critical for the smoothing theorem. Also, Proposition 3.10 is stated without proof, and Proposition 3.15(5) is explicitly deferred yet used in Example 4.2 to claim the deformation space has dimension 16 locally. Those are real dependencies, not just cosmetic gaps.\n\nStill, the paper deserves a serious referee. The framework is new, the examples are specific enough to check, and the main theorem is a plausible application of a published result. I'd send it to review and ask the referee to pin down Prop 2.2 or make the forward direction a clearly stated conjecture. The honest presentation of its own limitations makes me trust the author's judgment, even where the write-up is not complete.","headline":"Genuinely new framework for multiparameter degenerations with strong examples, but the smoothing theorem inherits risk from a sketched log-structure construction and two deferred results; worth a serious referee.","tokens_in":18117,"tokens_out":3578,"would_cite":true,"duration_ms":38046,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14D15","14D06","14J28","14J32"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that d-semistable, K-trivial 'polync' varieties—singular spaces locally modelled on products of normal-crossing singularities—have unobstructed log smooth deformations, giving multiparameter semistable smoothings with smoot","keywords":["polync varieties","d-semistability","log smooth deformations","Kulikov models","K3 surfaces","logarithmic Bogomolov–Tian–Todorov theorem","monodromy cone","semistable morphism"],"falsifier":"Run the omitted converse of Proposition 2.2 on a concrete S-colored polync variety, e.g. the rhombicuboctahedral polysnc K3 surface of Example 4.2: if its local transition systems cannot be chosen so that the product of units equals one on every double overlap, then no log structure of semistable type exists even though the L_s are trivial—disproving Proposition 2.2 and removing the input to the logarithmic Bogomolov–Tian–Todorov theorem that Theorem 2.5 requires.","tokens_in":17188,"feed_emoji":"🔷","tokens_out":6394,"duration_ms":105112,"temperature":0.7,"pith_summary":"This paper establishes a multiparameter generalization of the classical smoothability theorem for degenerations of Calabi–Yau varieties. It introduces 'polync varieties'—singularities that are locally products of normal-crossing singularities—and defines d-semistability for them via a color-coding of the dual polysimplicial complex. The central theorem states that any proper, d-semistable, colorable polync variety with trivial dualizing sheaf and algebraic components has unobstructed log smooth deformations, yielding a smoothing over a polydisk whose total space is smooth: a multiparameter Kulikov model. This matters because such smooth total spaces let one specialize algebraic cycle classes transversely to the central fiber, a tool recently used to attack questions about algebraic cycles and rationality. The paper also works out combinatorial period maps and monodromy cones for polysnc K3 surfaces, and gives explicit examples including a four-parameter model based on the rhombicuboctahedron and a ten-parameter degeneration of elliptic K3 surfaces.","feed_headline":"Polync degenerations smooth in several parameters at once","feed_subtitle":"Generalizing one-parameter normal-crossing smoothability, the proof yields explicit multiparameter Kulikov models.","key_machinery":"The load-bearing construction is the S-coloring of the dual polysimplicial complex of a polync variety: each simplex factor of each stratum gets a color, and stratum inclusions must be compatible. Colors index both the base parameters of the smoothing and the summands of the sheaf Ext^1(Ω^1, O), which turns out to be a direct sum of line bundles L_s supported on the color-s singular locus. d-semistability means each L_s is trivial; Proposition 2.2 converts this into a log structure of semistable type by gluing local transition systems. The logarithmic Bogomolov–Tian–Todorov theorem—applied to this log smooth, saturated, log Calabi–Yau object over the logarithmic point with monoid N^S—then de","core_discovery":"The paper claims that d-semistability—the vanishing, for each color, of an obstruction line bundle on the color's singular locus—is exactly the condition that a polync variety carries a 'log structure of semistable type.' Once that log structure is in hand, the logarithmic Bogomolov–Tian–Todorov theorem applies, because the log structure is log smooth and saturated and the dualizing sheaf is trivial. The conclusion is that log smooth deformations are unobstructed, and the universal formal deformation can be realized by an analytic semistable morphism from a smooth space to a polydisk of dimension equal to the number of colors plus the number of locally trivial parameters. The same machinery","pith_inferences":["Because the number of base parameters equals the number of colors, an S-coloring is not just bookkeeping: it is a combinatorial choice of how many smoothing directions are 'active', so one could design degenerations with prescribed monodromy cones by coloring dual complexes appropriately.","The same log-BTT route should extend to d-semistable polync degenerations of higher-dimensional Calabi-Yau and to abelian varieties via the Mumford construction, giving multiparameter Kulikov models beyond surfaces.","The period-map criterion for polysnc K3 surfaces suggests a direct combinatorial algorithm to decide smoothability of a given triangle-square decomposition of S^2, which could be automated and used to search for new explicit multiparameter degenerations.","If the omitted reverse direction of Proposition 2.2 fails in some multi-color case, one would still have the forward direction: log structures give d-semistability, so the smoothability theorem would apply anyway to the (possibly smaller) class of log-smoothable varieties."],"forward_implications":["The universal log smooth deformation of a d-semistable K-trivial polync variety is a Kulikov model over a polydisk, with smooth total space; any transverse S-dimensional slice is again a Kulikov model.","If the central fiber is projective with an ample line bundle, the deformation algebraizes to a projective semistable smoothing over a power series ring.","For polysnc Type III K3 surfaces, d-semistability is equivalent to a purely combinatorial period condition φ(ξ_G)=1 for every slab G, and the naive dimension of the d-semistable locally trivial deformation space is 20−|S|.","The monodromy cone is generated by vectors λ_s whose self and mutual intersections count triangles and squares of each color; for the paper's examples the Gram matrix is nondegenerate, giving a 10-dimensional nilpotent orbit recovering a known lattice-polarized K3 moduli space.","Polysnc surface Kulikov models subdivide into 1-parameter Kulikov models by crepant resolutions, preserving the dual complex's homeomorphism type."],"fun_headline_variants":["Multiparameter Kulikov models via d-semistability","Smooth polync degenerations now work in many parameters","d-semistable polyncs admit log smooth deformations","Generalizing Kulikov models to multiparameter bases","Smoothing K-trivial polync varieties in several directions"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The theorem stands on the equivalence in Proposition 2.2 between d-semistability and the existence of a log structure of semistable type; the reverse direction is omitted in the paper, and if it is false for some S-colored polync variety, the logarithmic Bogomolov–Tian–Todorov theorem can no longer be applied and the smoothability claim is unsupported.","fun_headline_variants_meta":{"raw":{"variants":["Multiparameter Kulikov models via d-semistability","Smooth polync degenerations now work in many parameters","d-semistable polyncs admit log smooth deformations","Generalizing Kulikov models to multiparameter bases","Smoothing K-trivial polync varieties in several directions"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000203,"raw_usage":{"total_tokens":1166,"prompt_tokens":632,"completion_tokens":534,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":376,"completion_tokens_details":{"reasoning_tokens":451}},"tokens_in":376,"tokens_out":534,"duration_ms":5831,"temperature":1.0,"reasoning_tokens":451,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T06:13:35.594861+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the omitted converse of Proposition 2.2 on a concrete S-colored polync variety, e.g. the rhombicuboctahedral polysnc K3 surface of Example 4.2: if its local transition systems cannot be chosen so that the product of units equals one on every double overlap, then no log structure of semistable type exists even though the L_s are trivial—disproving Proposition 2.2 and removing the input to the logarithmic Bogomolov–Tian–Todorov theorem that Theorem 2.5 requires.","supporting_citations":[],"review_version":1}