{"id":"7599b9a1-bd1b-4197-b3ab-ea7983b95908","arxiv_id":"2412.15913","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Very general noncommutative projective planes do not admit phantom categories.","lead":"The paper proves that very general noncommutative projective planes, built from a smooth elliptic curve and an infinite-order translation automorphism, have no phantom triangulated subcategories. It extends recent commutative-surface no-phantom theorems into the noncommutative setting and gives evidence for a broader conjecture.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Classification of autoequivalences sending skyscrapers to skyscrapers up to shift may be misapplied: spherical twists on elliptic curves are counterexamples, so the step T ≅ ρ_*(-⊗L)[n] is not justified.","rationale":"The paper is carefully written and most of the algebraic machinery is coherent: the spanning-class argument, the support-counting lemmas, and the spherical-functor framework are internally consistent. The reader correctly identified the triangle (2.19) and the classification of autoequivalences as potential weak points. My stress-test focuses on the latter, where there appears to be an actual gap. In the proof of Theorem 3.12, the classification of autoequivalences sending skyscrapers to skyscrapers up to shift is used to conclude that the spherical twist T is standard. But on a smooth elliptic curve, skyscraper sheaves are spherical objects, so spherical twists are nontrivial autoequivalences that preserve the set of skyscraper sheaves up to shift. The square of such a twist even preserves the K-class of every skyscraper, matching the identity [O_p] = [C_p] used in the proof, while not being standard. Therefore the inference from (3.23) to (3.24) is not justified by the cited result unless an additional argument shows the shift 2a_p is constant in p. The theorem may still be true, and the gap may be repairable, but as written the central argument does not close the case. For this reason the verdict should be conditional rather than unconditional acceptance.","tokens_in":16325,"tokens_out":62377,"duration_ms":544613,"concrete_test":"Verify the exact statement of [HB05, Corollary 4.3] and test the autoequivalence T = T_{O_p}^2 on a smooth elliptic curve E. Check that T(k(x)) is a shift of a skyscraper sheaf for every closed point x, that [T(k(x))] = [k(x)] in K0(E) for all x, but that T is not isomorphic to ρ_*(-⊗L)[n] for any automorphism ρ, line bundle L, and integer n. If this counterexample is valid, the step from (3.23) to (3.24) in the proof of Theorem 3.12 is invalid; alternatively, supply an argument proving that the shift 2a_p in (3.23) is independent of p and cite the fixed-shift version of the classification.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The proof of Theorem 3.12 hinges on the inference from (3.22)-(3.23) to (3.24): once C_p = T(O_p) is shown to be O_p[2a_p], the text cites [HB05, Corollary 4.3] to conclude that T is of the form ρ_*(-⊗L)[n]. This is the key step that later yields ρ = id, n = 0, L ≅ O_E, and finally that the kernel morphism id → T is an isomorphism. The cited classification is used in the form: an autoequivalence of D^b(E) sending skyscraper sheaves to skyscraper sheaves up to shift is standard. On a smooth elliptic curve, however, skyscraper sheaves are spherical objects (Ext^0 = Ext^1 = k, so RHom(O_p,O_p) ≅ k ⊕ k[-1]), and the spherical twist T_{O_p} is an autoequivalence of D^b(E). It sends O_q to O_q for q ≠ p and O_p to O_p[1]. Its square T_{O_p}^2 sends O_p to O_p[2] and fixes every other skyscraper, so it sends every skyscraper to a shift of a skyscraper and satisfies [T(O_q)] = [O_q] in K0(E) for all q, yet it is not of the form ρ_*(-⊗L)[n]. Thus the classification as stated in the paper either assumes a fixed shift independent of the point or does not allow shifts at all. The paper does not prove that the integer a_p in (3.23) is constant in p. Since the shift can vary (as the square of a spherical twist shows), the inference to (3.24) collapses, and with it the subsequent steps identifying T with the identity and deducing Lj^*B_p = 0 for every p.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves Theorem 1.3 (= Theorem 3.12): if A is a three-dimensional Artin–Schelter regular quadratic algebra associated to a geometric triple (E, σ, L), with E a nonsingular elliptic curve and σ a translation of infinite order, then the derived category D^b qgr(A) admits no phantom categories. In fact, the proof aims to show the stronger statement that every admissible subcategory B with K0(B) = 0 is trivial. The strategy follows the commutative arguments of Pirozhkov and Borisov–Kemboi: one proves that the objects {j^*O_p}_{p∈E} form a spanning class, studies the restricted functor Lj^*, uses the spherical twist associated to an admissible subcategory, and derives a contradiction from the existence of a point outside the countable exceptional set E_sp.","tokens_in":16718,"tokens_out":8307,"duration_ms":80267,"significance":"If the proof is correct, the result is a meaningful noncommutative analogue of the nonexistence of phantoms on del Pezzo surfaces and on certain non-generic blow-ups of P^2, and it gives evidence for Conjecture 1.4. The paper is clearly written, gives full background, and contains a substantial appendix (Proposition A.4) that supplies the dg-enhanced spherical-twist triangle needed in the main argument. The main theorem is also stronger than the stated phantom nonexistence, since it rules out all admissible subcategories with vanishing Grothendieck group. However, the proof as written contains a central gap in the classification step for the twist functor T, so the significance is currently conditional on repairing that step.","major_comments":[{"comment":"The inference from (3.23) to (3.24) is not justified. The paper applies [HB05, Corollary 4.3] to conclude that an autoequivalence T of D^b(E) sending every skyscraper sheaf to a shift of a skyscraper sheaf must be of the form ρ_*(-⊗L)[n]. The cited classification requires the shift to be independent of the point (or to be absent altogether), but the paper has only established C_p ≅ O_p[2a_p] with a_p possibly depending on p. This distinction is essential: on an elliptic curve, the square of the spherical twist at O_p is an autoequivalence satisfying T^2(O_p) ≅ O_p[2] and T^2(O_q) ≅ O_q for q≠p, so it sends skyscrapers to skyscrapers up to shift, preserves all K0 classes, and is not of the form ρ_*(-⊗L)[n]. Since the subsequent claims ρ = id, n = 0, L ≅ O_E, and the kernel morphism id → T being an isomorphism all rely on (3.24), Theorem 3.12 is not established unless constancy of a_p is proved or an alternative argument replacing this classification step is supplied.","section":"§3.3, Eq. (3.24)"}],"minor_comments":[{"comment":"The section title contains the typo “Maim theorem”; it should read “Main theorem”.","section":"Section 3 header"},{"comment":"“three dimentional” should be “three dimensional”.","section":"§2.2.1"},{"comment":"“quadchotomy” is not standard English; consider “four-way classification” or “quadrichotomy”.","section":"Eq. (2.16)"},{"comment":"“n ot” in the abstract should be “not”.","section":"Abstract"}],"recommendation":"major_revision","confidential_remarks":"The reader’s positive assessment of the overall structure is fair, but the skeptic’s objection lands: the classification step at (3.24) is a genuine gap, not a cosmetic one, because spherical twists on elliptic curves provide concrete autoequivalences satisfying the stated hypotheses with varying shifts. I would encourage the editor to request a revised version in which the author either proves that the shift a_p is constant under the additional assumptions of the theorem or replaces the argument with a different way to control T."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The stress-test concern is correct, and I read the proof more skeptically than the reader did. The paper proves a genuinely new, interesting theorem, but the argument as written has a load-bearing gap at the point where [HB05, Corollary 4.3] is used to conclude that the autoequivalence T is standard.\n\nWhat the paper does well: the adaptation of the Pirozhkov/Borisov-Kemboi strategy to noncommutative planes is not a mechanical translation. The spanning-class result (Prop. 3.3), the countability of the special set (Prop. 3.11), and the dg-enhancement work in the appendix are careful and go beyond simply quoting the commutative case. The paper is also honest about the limits of its methods, stating Conjectures 1.4 and 1.5 rather than overselling. The citation pattern looks fine; the external theorems are independent, and there is no circular reasoning.\n\nBut the step from (3.23) to (3.24) is not justified. On a smooth elliptic curve, the square of a spherical twist at a skyscraper sends every skyscraper to a shift of a skyscraper, with shift depending on the point (O_p goes to O_p[2], other points are fixed). It acts trivially on K0, so it satisfies all the conditions the paper has established up to that point. Yet it is not of the form ρ_*(-⊗L)[n]. The classification in [HB05] must either require a fixed shift across all points or exclude shifts altogether; the paper does not prove that the integer a_p in (3.23) is constant. The rest of the proof — identifying ρ, n, and L, and then showing the natural transformation id → T is an isomorphism — collapses if T is not standard. This is not a minor technicality; without this step, the conclusion Lj^*B_p = 0 for all p does not follow.\n\nThe underlying conjecture may well be true, and the gap might be fixable by proving the shift is constant or by using the full structure of autoequivalences of D^b(E) (generated by standard functors and spherical twists) and showing the twist T from the admissible subcategory cannot involve a spherical twist. But as written, the proof is incomplete.\n\nWho benefits from this paper: anyone working on derived categories of noncommutative surfaces, phantom categories, or the boundary between commutative and noncommutative rationality questions. It deserves a serious referee, because the question is important and the partial results are substantial, but I would not cite the main theorem as established until the gap is closed. The reader's ACCEPT verdict is too generous; this needs to go back to the author for a fix.","headline":"The main theorem is likely true, but the proof has a real gap at the step invoking [HB05, Cor 4.3]: varying shifts allow non-standard autoequivalences like spherical twists.","tokens_in":17206,"tokens_out":6945,"would_cite":false,"duration_ms":63324,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14F08","14H52","16S38","18G80"],"pacs":[],"model":"deepseek-v4-flash","headline":"Very general noncommutative projective planes admit no phantom categories.","keywords":["phantom category","noncommutative projective plane","Artin-Schelter regular algebra","spherical functor","derived category","elliptic curve","semiorthogonal decomposition"],"falsifier":"For a three-dimensional AS-regular algebra with an infinite-order translation, exhibit a nonzero admissible subcategory $\\mathcal{B}\\subset D^b\\mathrm{qgr}(A)$ with $K_0(\\mathcal{B})=0$; the theorem predicts none exists. A more local check is to compute the twist triangle (2.19) for the right orthogonal of a partial exceptional collection and find a point $p$ outside the countable exceptional set where the morphism $\\mathcal{O}_p \\to T(\\mathcal{O}_p)$ is zero, which would produce a nonzero $Lj^*\\mathrm{pr}^R_{\\mathcal{B}}j_*\\mathcal{O}_p$.","tokens_in":16135,"feed_emoji":"👻","tokens_out":12942,"duration_ms":102984,"temperature":0.7,"pith_summary":"The paper proves that very general noncommutative projective planes—flat deformations of the projective plane as an abelian category—carry no phantom categories. A phantom category is a nontrivial admissible subcategory of a derived category with trivial Grothendieck group and Hochschild homology. The proof actually establishes more: no admissible subcategory with vanishing Grothendieck group exists in $D^b\\mathrm{qgr}(A)$ for such a plane. This matters because phantom categories are known to exist on some rational surfaces, while the projective plane itself has none; the noncommutative setting is a new test of where phantoms can appear. The conclusion holds for the planes associated to a smooth elliptic curve and a translation of infinite order, which is the very general case.","feed_headline":"No phantom categories on very general noncommutative planes","feed_subtitle":"The proof rules out every admissible subcategory with trivial Grothendieck group on these noncommutative planes.","key_machinery":"The load-bearing object is the spherical restriction functor $Lj^*: D^b\\mathrm{qgr}(A) \\to D^b\\mathrm{coh}(E)$ from the noncommutative plane to its anti-canonical elliptic curve. The key identity is the exact triangle (2.19): for every admissible subcategory $\\mathcal{B}$, there is a triangle $Lj^*\\mathrm{pr}^R_{\\mathcal{B}} j_*F \\to F \\to T(F) \\to Lj^*\\mathrm{pr}^R_{\\mathcal{B}} j_*F[1]$, where $T$ is the spherical twist associated to the composition $\\mathcal{B} \\hookrightarrow D^b\\mathrm{qgr}(A) \\xrightarrow{Lj^*} D^b\\mathrm{coh}(E)$. Because $Lj^*$ is spherical, $T$ is an autoequivalence of $D^b\\mathrm{coh}(E)$, and the triangle compares the projection of $j_*F$ onto $\\mathcal{B}$ with the twist of $F$. The proof combines this triangle with the spanning class $\\{j_*\\mathcal{O}_p\\}$ and the countability of the exceptional support set to force $Lj^*\\mathrm{pr}^R_{\\mathcal{B}}j_*\\mathcal{O}_p=0$.","core_discovery":"The central result is Theorem 3.12: if $A$ is a three-dimensional AS-regular quadratic algebra coming from a geometric triple $(E,\\sigma,L)$ with $E$ a smooth elliptic curve and $\\sigma$ a translation of infinite order, then $D^b\\mathrm{qgr}(A)$ admits no phantom categories. In fact, for any semiorthogonal decomposition $D^b\\mathrm{qgr}(A)=\\langle \\mathcal{A},\\mathcal{B}\\rangle$ with $K_0(\\mathcal{B})=0$, the subcategory $\\mathcal{B}$ is zero. The proof shows that every object $j_*\\mathcal{O}_p$ lies in $\\mathcal{A}$ by feeding the spherical restriction functor $Lj^*$ into the twist triangle, using the classifications of spherical objects and autoequivalences on the elliptic curve to force the spherical twist to act trivially on skyscraper sheaves outside a countable exceptional set. Since $\\{j_*\\mathcal{O}_p\\}$ is a spanning class, this makes the right factor $\\mathcal{B}$ vanish.","pith_inferences":["The same two ingredients—a spanning class of skyscraper sheaves and countability of an exceptional support set—might be looked for in higher-dimensional noncommutative projective spaces; if they can be found, the no-phantom conclusion would likely extend there.","The theorem suggests that phantom categories on rational surfaces arise from commutative blow-up geometry rather than from noncommutative deformation; a natural test is to search for phantoms on blow-ups of noncommutative planes at points on the anti-canonical divisor.","Because the proof uses only $K_0(\\mathcal{B})=0$ and not the full phantom condition, the same argument may also rule out other 'small' admissible subcategories once their $K_0$ is torsion rather than zero, although the current triangle argument would need modification."],"forward_implications":["If the theorem is correct, then no phantom category can appear as a factor in any semiorthogonal decomposition of $D^b\\mathrm{qgr}(A)$ for these very general noncommutative projective planes.","The stronger statement rules out every admissible subcategory with trivial Grothendieck group, so the obstruction is not about Hochschild homology but about $K_0$ itself.","The result provides a noncommutative counterpart to the known absence of phantoms on the projective plane and certain rational surfaces, showing the no-phantom property is stable under very general noncommutative deformation.","The proof isolates a countable exceptional subset of the anti-canonical curve; all skyscraper sheaves outside this set are controlled by the twist argument, which is why the 'very general' qualifier is natural."],"supporting_citations":[{"why":"Supplies the GK-dimension and g-torsion results (Propositions 7.5 and 7.9) that force modules of bounded dimension to be g-torsion when sigma has infinite order; this underpins the spanning-class and support lemmas.","marker":"[ATV91]"},{"why":"Provides the classification of three-dimensional quadratic AS-regular algebras by geometric triples and the geometric algebra construction used throughout.","marker":"[ATV90]"},{"why":"Develops spherical functors and the twist/cotwist theory; its main theorems justify that the restriction functor is spherical and that admissible subcategories induce new spherical functors.","marker":"[Add16]"},{"why":"Gives the derived Morita theory for dg categories used in Appendix A to represent the projection and twist functors as integral kernels on $E\\times E$.","marker":"[To¨ e07]"},{"why":"Classifies spherical objects on elliptic curves as simple vector bundles or skyscraper sheaves up to shift, which converts the K-class equality into the isomorphism $C_p\\cong\\mathcal{O}_p[2a_p]$.","marker":"[BK06]"},{"why":"Classifies autoequivalences of an elliptic curve that send skyscraper sheaves to skyscraper sheaves up to shift, giving the form $T\\cong\\rho^*(-\\otimes L)[n]$.","marker":"[HB05]"},{"why":"Provides the commutative blow-up theorem whose strategy of spanning class plus support counting is adapted to the noncommutative setting.","marker":"[BK24]"},{"why":"Establishes the full strong exceptional collection $(\\mathcal{O},\\mathcal{O}(1),\\mathcal{O}(2))$ and the tilting object, giving $K_0\\cong\\mathbb{Z}^3$ and the dg setup needed in Appendix A.","marker":"[AOU14]"},{"why":"Computes the Serre functor of $D^b\\mathrm{qgr}(A)$ as $M\\mapsto M(-3)[2]$, which makes the restriction functor spherical and fixes the cotwist $C(M)=M(-3)[1]$.","marker":"[VdB97]"}],"fun_headline_variants":["No phantom categories on noncommutative projective planes","Phantom categories impossible on generic noncommutative planes","Very general noncommutative planes reject phantoms","Phantom categories banned on noncommutative planes","Noncommutative planes have no phantom categories"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof leans on the exact triangle (2.19), assembled in Appendix A from dg enhancements and a tilting object, and on the classifications of spherical objects and autoequivalences of the elliptic curve; if any of these structural inputs fails, the identification of the spherical twist with the identity on skyscraper sheaves would no longer follow.","fun_headline_variants_meta":{"raw":{"variants":["No phantom categories on noncommutative projective planes","Phantom categories impossible on generic noncommutative planes","Very general noncommutative planes reject phantoms","Phantom categories banned on noncommutative planes","Noncommutative planes have no phantom categories"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000388,"raw_usage":{"total_tokens":1952,"prompt_tokens":753,"completion_tokens":1199,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":369,"completion_tokens_details":{"reasoning_tokens":1122}},"tokens_in":369,"tokens_out":1199,"duration_ms":6433,"temperature":1.0,"reasoning_tokens":1122,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T11:00:39.372449+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For a three-dimensional AS-regular algebra with an infinite-order translation, exhibit a nonzero admissible subcategory $\\mathcal{B}\\subset D^b\\mathrm{qgr}(A)$ with $K_0(\\mathcal{B})=0$; the theorem predicts none exists. A more local check is to compute the twist triangle (2.19) for the right orthogonal of a partial exceptional collection and find a point $p$ outside the countable exceptional set where the morphism $\\mathcal{O}_p \\to T(\\mathcal{O}_p)$ is zero, which would produce a nonzero $Lj^*\\mathrm{pr}^R_{\\mathcal{B}}j_*\\mathcal{O}_p$.","supporting_citations":[],"review_version":1}