{"id":"a5b7d469-087b-4fd6-a183-41bdcbadf267","arxiv_id":"2607.19348","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Assuming Wahl's conjecture, every normal degeneration of the projective plane to a surface with only rational singularities is one of the Markov-equation family or one of six newly found surfaces.","lead":"This paper classifies all ways the complex projective plane can degenerate to a surface with only mild singularities, assuming a famous open conjecture about surface singularities. It finds six previously unknown examples and shows how they connect to known families.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 3.3's finite enumeration of QHD star-graph legs is asserted rather than demonstrated; a missed case would make Theorem 1.1's six-surface list incomplete even under Wahl's conjecture.","rationale":"The reader's CONDITIONAL verdict is appropriate. The paper is explicit that Theorem 1.1 assumes Wahl's conjecture, so the open conjecture is not an internal flaw; it is the stated hypothesis. The more pointed concern is internal: Lemma 3.3 is the combinatorial heart of the necessity direction, and its proof is a hand-check asserted in a few sentences without a transparent table matching Bhupal--Stipsicz types to leg forms. This is precisely a place where a missed case would silently invalidate the completeness claim. The concrete test proposed would settle it by an independent enumeration. I also noted a concrete arithmetical inconsistency in Lemma 2.4's displayed formula: as printed, the equation yields K^2_Wt + sum(mu_i) = 7 instead of 9, which would make the conclusion sum(mu_i)=0 unjustified. The subsequent display suggests the intended coefficient is 12, not 10, so this is almost certainly a typo rather than a mathematical error, but it should be fixed. The smoothing step's dependence on [CU, Thm. 5.1] with unstated hypotheses is a further reason to keep the verdict conditional. Since none of these observations decisively overturns the central claim, the reader's CONDITIONAL verdict stands unchanged; if the enumeration check found a counterexample, the verdict would move to REJECT, and if it confirms the list, ACCEPT would be justified.","tokens_in":10103,"tokens_out":16143,"duration_ms":160264,"concrete_test":"Write an independent script that enumerates all valency-3 QHD star graphs from [BS]/[SSW], computes their weighted dual graphs, and solves the continued-fraction condition [a_s,...,a_1,1,b_1,...,b_t]=0 together with the Section 2 constraints (d>=8, s,t>=1, at most one (-1)-curve in the fiber, and Lemma 3.2's exclusion). Then compare the solutions with the five type-(L1) cases (a,b,c,f,j) and the stated p,q,r values. If any additional solution appears, Theorem 1.1's list is incomplete; if exactly the listed cases appear, the enumeration is sound.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The completeness direction of Theorem 1.1 rests on Lemma 3.3, which reduces all possible valency-3 QHD star graphs from the Bhupal--Stipsicz classification to the four leg families (L1)--(L4) and then discards (L2)--(L4) by terse contradictions. The proof does not reproduce the relevant star graphs or provide a table matching each [BS] type to the leg data; it only states 'we have listed all possibilities for the leg [d, a_s,...,a_1]' and then gives one-line exclusions. A missing star-graph type, or a missing leg-length solution to the continued-fraction condition [a_s,...,a_1,1,b_1,...,b_t]=0 under the constraints of Section 2 (d>=8, s,t>=1, Lemma 3.2), would make the classification incomplete even if Wahl's conjecture is true. A separate concrete issue: the displayed formula in Lemma 2.4 reads 10*chi(O_X)-chi_top(X); with chi(O_X)=1 and b1(C_i)=0 this gives K^2_Wt + sum(mu_i) = 7, while the next display asserts 9. Replacing 10*chi by 12*chi reconciles the computation, so this is likely a typo, but it should be corrected and verified.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies normal degenerations of the complex projective plane with only rational singularities. Building on Bădescu, Manetti, and Hacking–Prokhorov, it proves (Theorem 1.1) that, assuming Wahl's conjecture that every QHD singularity is weighted homogeneous, every such degeneration is either a partial Q-Gorenstein smoothing of P(a^2,b^2,c^2) with a^2+b^2+c^2=3abc, or one of six explicitly listed surfaces W_j, W_f, W_c, W_c5, W_b, W_b13. The proof reduces the problem to a combinatorial analysis of the star graphs of weighted homogeneous QHD singularities via the Bhupal–Stipsicz classification, and constructs the six surfaces via Artin contractions and smoothings. The paper also analyzes the MMP over the new degenerations (Theorem 1.2) and gives a BPS-type example with a symplectic rational blow-down to P^2.","tokens_in":10417,"tokens_out":8710,"duration_ms":72357,"significance":"If correct, the main theorem completes the classification of normal degenerations of P^2 with rational singularities modulo Wahl's conjecture, merging the Markov-family degenerations of Manetti and Hacking–Prokhorov with six new non-log canonical examples. The new surfaces have singularities whose indices are not Markov numbers, and their existence shows that the log canonical assumption is essential for the earlier classifications. The paper is transparent about its conditional nature and provides a source for computational verification. However, the completeness proof contains a load-bearing enumeration that is asserted rather than fully documented, and the existence of the six smoothings depends on an unpublished same-author preprint; these points need to be addressed before the classification can be considered verified.","major_comments":[{"comment":"The completeness of the classification depends on this lemma, which asserts that the only possible legs are (L1)–(L4) and then discards (L2)–(L4). The proof does not reproduce the relevant star graphs from [BS] nor provide a table matching each graph type to the leg data. Statements such as 'we have listed all possibilities' and 'we have listed the corresponding duals' are not verifiable from the text. Since a missed case would make the six-surface list incomplete even under Wahl's conjecture, the authors should supply a full enumeration, e.g., an appendix or a reproducible computation, and give the explicit contradictions for each discarded type.","section":"Section 3, Lemma 3.3"},{"comment":"The displayed equation reads '10χ(O_X) − χ_top(X)'. With χ(O_X)=1 and the relation χ_top(X)=ρ(X)+2 (which is implicit in the subsequent computation), this gives K^2_Wt + ∑ μ_i = 7, not the claimed 9. The conclusion is consistent with replacing 10 by 12. This is likely a typo, but the equation is used to establish that the singularities are QHD. Please correct it and verify the Noether-theoretic derivation.","section":"Section 2, Lemma 2.4"},{"comment":"The global existence of a smoothing for the six surfaces is imported from [CU, Thm. 5.1] and [CU, Rem. 5.2], a same-author preprint. The paper should state the hypotheses of the quoted theorem and justify that they hold for the constructed surfaces, or provide a self-contained proof. Since the theorem's 'exactly these six' claim requires that each listed surface actually appears as a degeneration, this dependency is load-bearing. The reader cannot currently verify this step from the manuscript alone.","section":"Proof of Theorem 1.1"}],"minor_comments":[{"comment":"The continued-fraction notation [a_s,...,a_1,1,b_1,...,b_t] and the term 'leg' should be defined before Lemma 3.1, as they are central to the enumeration.","section":"Section 3"},{"comment":"The star graphs from [BS] that are used in Lemma 3.3 are not reproduced. At least the specific types invoked in the proof should be displayed or referenced with precise labels so the reader can follow the 'only possible' claims.","section":"Figures 2 and 3"},{"comment":"The proof is very terse: 'For (L4), the only possible is (g)' and similar assertions are not substantiated. A table of QHD types with their leg forms and the corresponding duals would greatly improve readability and verifiability.","section":"Section 3, Lemma 3.3 proof"},{"comment":"The phrase 'one checks by the duality of continued fractions' appears twice; please expand these checks or provide a reference.","section":"Proof of Theorem 1.1"},{"comment":"The list of Γ^-·K_W' values and the flipped-curve data are asserted as 'straightforward computations'. Providing at least one sample computation would help the reader verify the MMP claims.","section":"Section 4, Lemma 4.1 and Theorem 1.2"},{"comment":"The labels in the figure do not match the notation in the text (e.g., 'W_f8' appears instead of W_f, 'W_as-8' instead of W_c). Please harmonize the figure labels with the theorem statements.","section":"Figure 1"},{"comment":"The phrase 'Suppose that some of these examples satisfy Remark 2.1' is vague; specify which hypothesis is needed for the BPS construction and why the particular example k=1, n=7 satisfies it.","section":"Section 5"},{"comment":"References [B2], [BPS], [CU], and [C] are listed with year 2026 and no publication venue; mark them as preprints or in-preparation.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper is a serious contribution and the conditional result is plausibly correct, but the proof of Lemma 3.3 is a genuine gap in the completeness argument, not just a presentation issue. The dependence on the same-author preprint [CU] for a load-bearing step also needs clarification. If the authors can provide a complete, verifiable enumeration and state the hypotheses of the quoted smoothing theorem, I would be inclined to accept the paper after revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Markus – Worth a proper look. The paper completes the Manetti/Hacking–Prokhorov classification of normal degenerations of P^2 with rational singularities, conditional on Wahl's conjecture (every QHD singularity is weighted homogeneous). That conditional framing is stated plainly, which I respect. Under that assumption, they prove the degeneration is either in the Markov-equation family or is one of six surfaces with non-log canonical QHD singularities. The four new singularity types are new to degenerations of P^2, and the BPS example in Section 5 is a nice separate spin-off.\n\nThe actual work: they reduce to a finite enumeration of star graphs from Bhupal–Stipsicz, prove structural lemmas (Lemma 2.5, Corollary 2.6) that pin down the Hirzebruch surface and number of fibers, then run the enumeration and construct the six surfaces via Artin contraction. They also show the smoothing exists via [CU, Thm. 5.1] and that the central fiber is P^2. The material in Section 4 (flips, MMP) is more intricate and I didn't re-verify the continued fraction lists, but the structure is coherent.\n\nSoft spots, in order:\n\n1. Lemma 3.3 is the load-bearing completeness step. It asserts that all possible QHD valency-3 star graphs reduce to the four leg families (L1)–(L4), and then rules out (L2)–(L4) in a few lines. The text does not reproduce the Bhupal–Stipsicz star graphs or give a table matching each type to the leg data. For a paper whose main theorem is a classification, that is a real gap in exposition. It may be fixable with a table, but a referee should ask for it.\n\n2. Lemma 2.4 has a numerical inconsistency: the displayed formula uses 10*chi(O_X), but the following equality gives K^2 + sum mu = 9, which works if you use 12*chi(O_X). Likely a typo (Noether formula coefficient), but it sits right next to the claim mu_i = 0, so it should be fixed before publication.\n\n3. The proof imports several results from the same-authors' companion paper [CU] without restating hypotheses. In particular, [CU, Thm. 5.1] carries the global smoothing existence. If that theorem has hidden assumptions, the construction of the six degenerations weakens. The paper points to it, but since the main theorem rests on it, a referee should check those hypotheses carefully.\n\n4. The whole theorem is conditional on Wahl's conjecture, so the title's claim is a classification modulo that conjecture. That is stated honestly in the abstract and intro, but it means the enumeration in Lemma 3.3 only covers weighted homogeneous QHD singularities. If the conjecture fails, the classification could miss non-weighted-homogeneous cases. Not a flaw in the paper's logic, but the strength of the theorem is tied to an open problem.\n\nOverall: this is a serious, well-organized paper with a clear main theorem and, as far as I could verify, no internal contradiction. The enumeration gap is the thing I'd want pinned down before trusting the completeness direction. The typo is minor. I'd send it to a good referee rather than desk-reject, and I'd bring it to the reading group if anyone cares about P^2 degenerations or QHD singularities.","headline":"A conditional but honest completion of the P^2 degeneration classification: six genuinely new non-log canonical degenerations, one enumeration gap worth checking, and a clear reliance on Wahl's conjecture.","tokens_in":10917,"tokens_out":2280,"would_cite":true,"duration_ms":21963,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14J17","14J26","14D15"],"pacs":[],"model":"deepseek-v4-flash","headline":"Assuming the QHD weighted-homogeneity conjecture, every normal degeneration of P2 with only rational singularities is either a partial Q-Gorenstein smoothing of P(a²,b²,c²) with a²+b²+c²=3abc or one of six explicitly listed surfaces.","keywords":["degenerations of the complex projective plane","rational singularities","rational homology disk smoothings","weighted homogeneous singularities","QHD conjecture","Q-Gorenstein smoothing","non-log canonical surfaces","symplectic rational blow-down"],"falsifier":"A concrete way to test the theorem: search for a normal degeneration of P2 whose special fiber has only rational singularities and whose unique non-quotient singularity is either not weighted homogeneous or has a minimal-resolution star graph of valency four. The paper proves valency four cannot occur and enumerates all weighted homogeneous QHD graphs of valency three; any such degeneration would disprove the classification. Since the proof begins by showing every singularity in W has Milnor number zero, a degeneration whose special fiber contains a rational singularity that is provably not QH","tokens_in":9995,"feed_emoji":"📐","tokens_out":8049,"duration_ms":74255,"temperature":0.7,"pith_summary":"This paper aims to finish the classification of all ways the complex projective plane can degenerate to a normal surface with only rational singularities. Under the conjecture that every singularity admitting a rational-homology-disk smoothing is weighted homogeneous, it proves that the only such degenerations are the known partial Q-Gorenstein smoothings of weighted projective planes P(a²,b²,c²) with a²+b²+c²=3abc, together with six new surfaces (Wj, Wf, Wc, Wc5, Wb, Wb13). Four of the new surfaces carry a single non-log-canonical singularity; the other two carry two singularities and arise by slidings from the first four. The paper also shows that flipping the new degenerations in the minimal model program produces known Q-Gorenstein degenerations of P2 (or of weighted projective planes), and constructs a non-smoothable normal surface whose symplectic rational blow-down is P2. A reader should care because, if the conjecture holds, the enumeration of such degenerations is now closed.","feed_headline":"Six new degenerations complete the P2 list, pending QHD conjecture","feed_subtitle":"Assuming a QHD conjecture, every rational degeneration of P2 is either a weighted projective plane from a²+b²+c²=3abc or one of six new surf","key_machinery":"The engine of the proof is the star dual graph of the minimal resolution of the unique non-quotient singularity of W, sitting inside a blow-up of a rational ruled surface Fd. A count forces exactly one special fiber, a distinguished 'contractible leg,' and d = 6 + s1 - t1 ≥ 8. The possible legs are then enumerated by the continued-fraction identity [as,...,a1,1,b1,...,bt]=0; after checking each QHD star type, only five configurations remain (types (a), (b), (c), (f), (j)), four of which are impossible by the contractible-leg constraint, leaving the six surfaces. The contractibility criterion makes the combinatorics geometric, and a companion no-obstruction theorem globalizes the local QHD sm","core_discovery":"The central discovery is a complete list, conditional on the weighted-homogeneity conjecture: each degeneration (W ⊂ W) over a disk with general fiber P2 and special fiber W having only rational singularities must belong to one of seven families. Six are new and have one or two singularities; their minimal resolutions consist of chains that blow up a rational ruled surface F8 and then contract to yield four previously unseen non-log-canonical singularity types, with indices 58, 9, 8, and 16. The argument shows that every singularity of such a W is a rational-homology-disk (QHD) singularity, because the Milnor number of the induced smoothing at each point is forced to be zero by the rationali","pith_inferences":["Inference: The six-surface list likely survives if the weighted-homogeneity conjecture is replaced by a broader classification, because the construction only needs the star graphs to be QHD; the main risk is that non-weighted-homogeneous QHD singularities might create additional, currently invisible degenerations.","Inference: The paper's MMP analysis hints that the new non-log-canonical degenerations are flips away from log-canonical ones; iterating flips may produce a finite graph of degenerations of P2 connected by birational transformations, which could be worth mapping explicitly.","Inference: A natural testable extension is to search symplectically for analogues of the four new singularities: if each admits a rational-homology-disk Stein filling but not a QHD smoothing, they would form a symplectic family parallel to the complex classification."],"forward_implications":["If the weighted-homogeneity conjecture is true, the classification of normal degenerations of P2 with only rational singularities is closed: every example is either a partial Q-Gorenstein smoothing of a weighted projective plane P(a²,b²,c²) with a²+b²+c²=3abc or one of the six new surfaces.","The four new singularity types (indices 58, 9, 8, 16) show that indices of such degenerations need not be Markov numbers, refining expectations about which quotient singularities can arise.","Type (A2) degenerations contain at most three singularities; the new examples attain one or two.","Each of the six new degenerations, after a semistable flip, maps to a known Q-Gorenstein degeneration: P2, or the weighted projective planes 1/52(1,4) and 1/132(1,25), so the new family is connected to the old one through the minimal model program.","The constructed non-smoothable surface shows a singularity with a Stein rational-homology-disk filling but no QHD smoothing whose symplectic rational blow-down is P2; this provides a symplectic counterpart in this setting."],"fun_headline_variants":["Six new P2 degenerations join Markov-classified family","P2 degenerations fully classified, assuming Wahl's conjecture","Six new surfaces complete P2 degeneration list, conditional on QHD","Rational singularities: P2 degenerations classified, six new found","New degenerations of P2 with rational singularities: six added"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The classification depends on a long-open conjecture: any surface singularity that can be smoothed with a rational homology disk must be weighted homogeneous; if a counterexample exists, the star-graph enumeration could miss legitimate QHD singularities and the six-surface list might not exhaust all degenerations.","fun_headline_variants_meta":{"raw":{"variants":["Six new P2 degenerations join Markov-classified family","P2 degenerations fully classified, assuming Wahl's conjecture","Six new surfaces complete P2 degeneration list, conditional on QHD","Rational singularities: P2 degenerations classified, six new found","New degenerations of P2 with rational singularities: six added"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000224,"raw_usage":{"total_tokens":1228,"prompt_tokens":606,"completion_tokens":622,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":350,"completion_tokens_details":{"reasoning_tokens":535}},"tokens_in":350,"tokens_out":622,"duration_ms":6106,"temperature":1.0,"reasoning_tokens":535,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T12:39:45.383433+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete way to test the theorem: search for a normal degeneration of P2 whose special fiber has only rational singularities and whose unique non-quotient singularity is either not weighted homogeneous or has a minimal-resolution star graph of valency four. The paper proves valency four cannot occur and enumerates all weighted homogeneous QHD graphs of valency three; any such degeneration would disprove the classification. Since the proof begins by showing every singularity in W has Milnor number zero, a degeneration whose special fiber contains a rational singularity that is provably not QH","supporting_citations":[],"review_version":1}