{"id":"f0509605-5917-4bf8-9c9a-cded9ada9992","arxiv_id":"2412.08531","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Invariant stability conditions on local Calabi-Yau threefolds are shown to be fixed points of a group action, with all Donaldson-Thomas invariants computed explicitly.","lead":"This paper constructs special stability conditions on local Calabi-Yau threefolds such as the canonical bundle of P^1×P^1, and computes all associated Donaldson-Thomas invariants. The result explains why these stability conditions exist, and gives the first analytic wall-crossing structures for such threefolds with compact divisors.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Triviality of the Z2-action on Stab0(Y) is load-bearing; a nontrivial deck transformation would shrink the invariant locus and change the DT spectrum.","rationale":"The reader's weakest assumption is exactly the point on which the whole construction turns. If the H-action were a nontrivial deck transformation of the covering ϖ, the fixed locus Stab0(Y)^H would be a union of lower-dimensional strata, and the image under F would not be a connected component of Stab(X)^G; the classes and DT invariants of Theorem 4.3 would be those of the smaller locus, not the full component. I agree this is the load-bearing premise. However, the concern is readily testable and, by the argument in Section 4.2, already essentially established: the nontrivial element g of H fixes each stable object OC(n)⊗χ_p up to isomorphism (since g is the identity on the zero section C) and maps skyscrapers to skyscrapers, so the per-phase semistable subcategories are preserved; because the central charge is unchanged, the whole slicing is fixed. Since Stab0(Y) is a regular covering, a deck transformation fixing one point is the identity, so the action is trivial on the component. I would not change the reader's ACCEPT, but the concrete check above would settle the residual doubt. No other issue appears to threaten the central claim.","tokens_in":17407,"tokens_out":36819,"duration_ms":375756,"concrete_test":"Choose a reference point σ0 ∈ Stab0(Y) with heart Coh(Y) and central charge Z(β) = −1, Z(δ) = i. Let g be the nontrivial element of H = Z2 acting on Y by (z1,z2) ↦ (−z1,−z2) on fibers. Verify explicitly that g_*σ0 = σ0: (1) check g_*Coh(Y) = Coh(Y) as subcategories; (2) check that for each n, g_*(OC(n) ⊗ V) ≅ OC(n) ⊗ V, and that g_* maps the full subcategory of zero-dimensional sheaves to itself; (3) confirm that Z(g_*E) = Z(E) for all E. Since deck transformations of a regular covering act freely on the fiber, this implies the deck transformation is the identity, so H acts trivially on the whole component Stab0(Y). If instead g_*σ0 lies in the same fiber but is not equal to σ0, the invariant locus is proper and Theorem 4.3 requires revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim Theorem 4.3 rests on the identification Stab0(X)^G ≅ Stab0(Y), which requires that the Z2 action on the resolved conifold stability space Stab0(Y) is trivial. Section 4.2 asserts: 'The group H acts trivially on Γ = H∗(Y,Z) ... This implies that the induced action of H on Stab0(Y) is trivial.' Acting trivially on Γ is not by itself sufficient: the map ϖ: Stab0(Y) → {Z(β)≠0, Z(δ)≠0} is a covering with nontrivial deck transformations (e.g., tensoring with line bundles pulled back from the base acts trivially on Γ but permutes the OC(n) rays). Thus the argument must use the additional fact, stated in the same sentence, that H preserves the class of semistable objects listed in Theorem 4.1. That fact is plausible—the nontrivial element fixes each OC(n) up to isomorphism and maps skyscrapers to skyscrapers—but if H acted by a nontrivial deck transformation, Stab(Y)^H would be a proper subset, and the image Stab0(X)^G would not be a full connected component; the ray spectrum and the DT invariants in Theorem 4.3 would change. The paper's proof of triviality is a single sentence, so this premise deserves explicit verification.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper constructs invariant stability conditions on the derived category of the local Calabi-Yau threefold X = ω_{P^1×P^1} by exploiting an action of H = Z/2 on the resolved conifold Y = O_{P^1}(-1)^{⊕2}. Using the derived McKay correspondence and Dell's inducing theorem, it identifies a connected component of the fixed locus Stab(X)^G with the standard component Stab_0(Y) of the conifold stability space. The authors then describe all semistable objects on this locus, compute the associated Donaldson-Thomas invariants (Theorem 4.3), and conclude that the resulting DT theory defines analytic wall-crossing structures in the sense of Kontsevich-Soibelman (Theorem 1.2). An algebraic parallel via quivers with potentials is developed, and further examples (pseudo-dP_5, dP_3, and Y^{N,0} geometries) are discussed.","tokens_in":17678,"tokens_out":25034,"duration_ms":235984,"significance":"If the results hold, the paper gives a complete and explicit description of the BPS spectrum on the invariant stratum of stability conditions for a local Calabi-Yau threefold containing compact divisors, and provides the first known instances of analytic wall-crossing structures in this setting. The main theorems are reductions to established results—Dell's inducing theorem, Toda's description of the conifold, and Bryan-Cadman-Young's orbifold vertex—so the proofs are mostly reliable. The paper also clarifies the origin of the 'fine-tuned' stability conditions previously found in the physics and mathematics literature. The presentation is clear, and the explicit DT invariants are concrete, falsifiable statements. The main weaknesses are a terse proof of a load-bearing identification and a likely index-range typo in the Y^{N,0} theorem.","major_comments":[{"comment":"The condition \"0 < a ≤ b < N−1\" in the statement of Theorem 5.1 excludes the single-variable terms needed for N=2. As written, for N=2 the middle line is empty, so the theorem omits the invariants Ω((γ1+γ2)+nδ) = −2 and Ω((γ3+γ4)+nδ) = −2 that appear in Theorem 4.3; this contradicts the claim in the text that \"for N=2 it reproduces the DT invariants for the Z2-invariant stability conditions on ω_{P1×P1}\". The generating function (26) uses the range 0 < a ≤ b < N, which is the correct condition. The theorem should be corrected to 0 < a ≤ b < N (equivalently b ≤ N−1).","section":"Section 5.3, Theorem 5.1"},{"comment":"The identification Stab0(X)^G ≅ Stab0(Y) is load-bearing for the description of semistable objects and for the DT invariant computation of Theorem 4.3. The proof is compressed into the sentence: \"The group H acts trivially on Γ = H*(Y,Z) and preserves the class of semistable objects listed in Theorem 4.1. This implies that the induced action of H on Stab0(Y) is trivial.\" This implication is not fully justified: one must show that the autoequivalence h_* induced by the H-action fixes each stability condition σ ∈ Stab0(Y), not merely that it preserves the set of isomorphism classes of stable objects. Since h_* acts trivially on Γ, the central charge is fixed, and because h_* preserves the torsion pair generated by the OC(n) and the zero-dimensional sheaves, it preserves the heart and the slicing. Please expand this argument to make the triviality of the H-action explicit.","section":"Section 4.2"}],"minor_comments":[{"comment":"The proof refers to equation (13) when checking that the bijection preserves quasi-numerical stability conditions, but no equation (13) is displayed; the reference should be to the continuity bijection obtained a few lines above.","section":"Theorem 2.2 proof"},{"comment":"The conversion from the orbifold vertex formula to the stated DT invariants is delegated to [24, Sections 7.5.2–7.5.4]. A few lines spelling out how the exponents -2m and -4m translate into the invariants Ω((γ1+γ2)+nδ)=-2 and Ω(nδ)=-4 would improve readability and verifiability.","section":"Section 4.3, proof of Theorem 4.3"},{"comment":"The displayed computation of W_{PdP5/Z2} contains an intermediate term with a duplicated factor \"-X(2)1,3 X(2)3,5 X(2)5,7 X(2)7,1\" appearing twice; the final expression is correct, but the intermediate line should be cleaned up.","section":"Section 5.1, equation (17)"},{"comment":"The sentence \"The stable objects for the corresponding invariant stability conditions on Dc(ωdP3).\" is incomplete; it appears to be a leftover fragment and should be completed or removed.","section":"Section 5.2"},{"comment":"The ray labels use an integer k that is not defined in the caption; relating k to the integer n in Theorem 4.3 would help the reader connect the figure to the theorem.","section":"Figure 2"}],"recommendation":"major_revision","confidential_remarks":"The paper is well within the scope of the journal and the main ideas are sound. The major issues are local and fixable: a clear typo in the range of indices in Theorem 5.1, and a need for a more detailed justification of the triviality of the H-action on Stab0(Y). Neither issue undermines the central example, but both should be addressed before publication. The references to prior work appear appropriate, and the novelty with respect to [15,16,31] is clearly stated."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Hi,\n\nHere's the take on Bridgeland–Del Monte–Giovenzana, arXiv:2412.08531.\n\nThis is a genuinely useful note. The main content: they explain the 'fine-tuned' stability conditions on local P1×P1 (previously constructed by Xiong and suggested by physicists) as fixed points of a Z2 action, via Dell's inducing result. They then compute the full DT spectrum on that invariant locus (Theorem 4.3) and, using a result of Kontsevich–Soibelman, bootstrap to an analytic wall-crossing structure on a whole connected component of Stab(X). The Y^{N,0} family and the dP3 examples fill out the picture. I think the central claims hold up.\n\nWhat's genuinely new: the fixed-locus interpretation, the complete DT invariants, and the analytic wall-crossing application. The paper is honest about provenance: the stability conditions themselves are due to Xiong, and the DT computation on the central ray imports Bryan–Cadman–Young's orbifold vertex. The main theorems reduce to existing results (Dell, Toda, BCY) in a clean and checkable way.\n\nSoft spots: reliance on Dell's unpublished preprint [17] is a real dependency; if that stays in preprint limbo, the foundations are only as solid as Dell's theorem. The referee should look at that carefully. Also, Theorem 5.1 is proved by 'essentially the same argument' as Theorem 4.3, which is fine but somewhat condensed; the dP3 section is explicitly a sketch, with no DT invariants computed. These are minor issues for a note like this.\n\nOne thing I want to flag because the reading-group note raised it: the concern about the Z2 action on Stab0(Y) possibly being a nontrivial deck transformation doesn't hold up. The paper states that H acts trivially on Γ and—crucially—preserves the explicit class of semistable objects listed in Theorem 4.1. Since phases are determined by the central charge, which is unchanged, each phase subcategory is preserved, so the action is trivial on the whole component. The one-sentence proof is terse, but the ingredients are all there. I'd ask the authors to expand that into two sentences for the reader's sake, but it is not a gap.\n\nBottom line: this is a solid paper for people working on stability conditions, DT theory, and BPS structures. It deserves a serious referee and likely acceptance with minor revisions. I'd cite it, and I'd take it to a reading group.\n\nBest,\n\n[Your name]","headline":"Solid and honest note: the invariant stability conditions are explained as a fixed locus, the full DT spectrum on that locus is computed, and the stress-test worry about a nontrivial deck transformation on Stab0(Y) does not survive a careful reading.","tokens_in":18253,"tokens_out":5876,"would_cite":true,"duration_ms":62767,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14J32","14F05","14N35","16G20"],"pacs":[],"model":"deepseek-v4-flash","headline":"On the local Calabi-Yau threefold $X=\\omega_{\\mathbb{P}^1\\times\\mathbb{P}^1}$, the paper proves that a $\\mathbb{Z}_2$-invariant locus in the stability space has an explicitly enumerated stable spectrum, computes all nonzero…","keywords":["stability conditions","Donaldson-Thomas invariants","BPS spectrum","local Calabi-Yau threefolds","quivers with potential","resolved conifold","analytic wall-crossing structures","brane tilings"],"falsifier":"A direct check would be to compute the induced $\\mathbb{Z}_2$ action on the covering $\\varpi:\\mathrm{Stab}_0(Y)\\to\\mathrm{Hom}_{\\mathbb{Z}}(\\Gamma,\\mathbb{C})$ and on the Hom-spaces among the line bundles $\\mathcal{O}_C(n)$ on the resolved conifold; if any deck transformation is nontrivial, or if the action sends the standard heart to a different heart, then the fixed locus is a proper subset of $\\mathrm{Stab}_0(Y)$ and the identification used for Theorem 4.3 fails.","tokens_in":17191,"feed_emoji":"🧮","tokens_out":14970,"duration_ms":138110,"temperature":0.7,"pith_summary":"The paper addresses a question raised by physics computations: why do some local Calabi-Yau threefolds admit stability conditions with an unexpectedly simple BPS spectrum? For the total space of the canonical bundle over $\\mathbb{P}^1\\times\\mathbb{P}^1$, it answers by showing that these special stability conditions are exactly the ones fixed by a $\\mathbb{Z}_2$ symmetry of the derived category, and that the fixed locus is isomorphic to the well-understood stability space of the resolved conifold. On this invariant locus every stable object is explicit, and the paper computes the complete set of nonzero Donaldson-Thomas invariants, which fill just four families of rays (Theorem 4.3). The boundedness of these invariants lets the paper conclude that the Donaldson-Thomas theory on the whole connected component defines an analytic wall-crossing structure, which it presents as the first such structure for a local Calabi-Yau threefold with compact divisors. The same induction is applied to the $Y^{N,0}$ family of toric threefolds, where analogous explicit formulas hold.","feed_headline":"Four formulas give every DT invariant on an invariant locus","feed_subtitle":"On the P1×P1 canonical bundle, four simple formulas enumerate the BPS spectrum and imply analytic wall-crossing.","key_machinery":"The mechanism is the induction of stability conditions along a quotient, packaged as an isomorphism of fixed loci: given a finite abelian group $H$ acting on a threefold $Y$ with locally trivial canonical bundle, the crepant resolution $X\\to Y/H$ and the associated derived equivalence $D_c(X)\\cong D_c^H(Y)$ produce an isomorphism $\\mathrm{Stab}(X)^G\\cong \\mathrm{Stab}(Y)^H$, where $G$ is the character group of $H$, and an object is semistable on the $X$-side exactly when its image under the forgetful functor is semistable on the $Y$-side. In the main example $Y$ is the resolved conifold $\\mathcal{O}_{\\mathbb{P}^1}(-1)^{\\oplus 2}$, $H=\\mathbb{Z}_2$ acts by $-1$ on the fibres, and $X=\\omega_{\\mathbb{P}^1\\times\\mathbb{P}^1}$ is the crepant resolution of the quotient. Because $H$ acts trivially on the standard connected component $\\mathrm{Stab}_0(Y)$, the invariant locus in $\\mathrm{Stab}(X)$ is identified with the entire conifold component. The same induction is also realized algebraically by a $\\mathbb{Z}_2$ symmetry of the four-vertex quiver with potential whose quotient is the two-vertex conifold quiver, giving a second route to the same invariant stability conditions and connecting the construction to brane tilings.","core_discovery":"The central claim is Theorem 4.3. For an invariant stability condition $\\sigma$ on $D_c(X)$, $X=\\omega_{\\mathbb{P}^1\\times\\mathbb{P}^1}$ (equivalently, on the Jacobi algebra of the four-vertex quiver with potential shown in Figure 1(a)), the complete list of nonzero Donaldson-Thomas invariants is $\\Omega((n+1)\\gamma_1+n\\gamma_2)=\\Omega(n\\gamma_1+(n+1)\\gamma_2)=1$, $\\Omega((n+1)\\gamma_3+n\\gamma_4)=\\Omega(n\\gamma_3+(n+1)\\gamma_4)=1$ for all $n\\in\\mathbb{Z}$, $\\Omega((\\gamma_1+\\gamma_2)+n\\delta)=\\Omega((\\gamma_3+\\gamma_4)+n\\delta)=-2$ for all $n\\in\\mathbb{Z}$, and $\\Omega(n\\delta)=-4$ for all $n\\in\\mathbb{Z}\\setminus\\{0\\}$, with $\\delta=\\gamma_1+\\gamma_2+\\gamma_3+\\gamma_4$. The non-central rays each contain two spherical stable objects with no extensions between them, giving the $+1$ values; the central ray consists of zero-dimensional equivariant sheaves on the resolved conifold orbifold, and the $-2$ and $-4$ values are read from the orbifold vertex expansion [11]. From this explicit bounded spectrum, the paper concludes that the Donaldson-Thomas theory on the whole connected component containing these points defines an analytic wall-crossing structure in the sense of [25], and that the associated DT Riemann-Hilbert problem has the trivial solution $X(\\gamma)=\\exp(Z(\\gamma)/\\epsilon)$.","pith_inferences":["The same induction should apply to any local Calabi-Yau threefold that is a crepant resolution of a finite quotient of the resolved conifold; the $Y^{N,0}$ and pseudo-$dP_5$ examples then look like members of a general pattern rather than isolated coincidences.","The trivial Riemann-Hilbert solution at the invariant points, combined with the paper's observation that nontrivial solutions appear on a nearby codimension-one locus, suggests reading the invariant stratum as the 'algebraic' locus of a cluster integrable system; the paper does not explicitly draw this conclusion.","A testable extension is to deform the surface away from $\\mathbb{P}^1\\times\\mathbb{P}^1$ while preserving the quotient-quiver symmetry and check whether the fixed-locus identification with the conifold component persists; if it does, the phenomenon is robust under deformation."],"forward_implications":["For $X=\\omega_{\\mathbb{P}^1\\times\\mathbb{P}^1}$, the DT invariants of every stability condition in the connected component containing the invariant locus are determined in principle by Theorem 4.3 together with the wall-crossing formula, so the whole BPS spectrum of that component is under control.","Every invariant stability condition yields a convergent, integral BPS structure, and its DT Riemann-Hilbert problem has the trivial explicit solution $X(\\gamma)=\\exp(Z(\\gamma)/\\epsilon)$.","The connected component carries an analytic wall-crossing structure, which the paper states is the first known instance for a local Calabi-Yau threefold with compact divisors.","The same method produces explicit DT invariants for the infinite family $Y^{N,0}$: $\\Omega(\\gamma_{2j-1}+nv_j)=\\Omega(\\gamma_{2j}+nv_j)=1$, $\\Omega(\\pm\\sum_{j=a}^b v_j+n\\delta)=-2$, and $\\Omega(n\\delta)=-2N$.","The dP3 and pseudo-dP5 examples give further invariant loci whose quotient quivers are explicitly described, so their semistable objects and DT invariants can in principle be computed by the same mechanism."],"supporting_citations":[{"why":"Supplies the inducing/fixed-locus theorem that identifies $\\mathrm{Stab}(X)^G$ with $\\mathrm{Stab}(Y)^H$ and detects semistability via the functor $\\Theta$; this is the engine of both the geometric and algebraic constructions.","marker":"[17]"},{"why":"Provides the complete description of the standard connected component of the resolved conifold stability space, including its stable objects and covering of central-charge space; the invariant locus is identified with this component.","marker":"[29]"},{"why":"Gives the derived equivalence $D_c(X)\\cong D_c^H(Y)$ for the crepant resolution of a quotient, which transfers the $G$-action and the stability conditions to the local threefold.","marker":"[9]"},{"why":"Computes the orbifold ideal-sheaf Donaldson-Thomas invariants for zero-dimensional equivariant sheaves on the resolved conifold orbifold, giving the $-2$ and $-4$ values on the central ray.","marker":"[11]"},{"why":"Supplies the growth-rate criterion used to conclude, from the explicitly bounded spectrum of Theorem 4.3, that the DT theory on the connected component defines an analytic wall-crossing structure.","marker":"[25]"}],"fun_headline_variants":["Four formulas list all DT invariants on P1xP1 CY3","Complete BPS spectrum on the P1xP1 canonical bundle","Explicit DT invariants imply trivial wall-crossing on a CY3","All DT invariants on a local Calabi-Yau from quiver data"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the paper's assertion that the $\\mathbb{Z}_2$ symmetry fixes every stability condition in the standard component of the resolved-conifold stability space, making the invariant locus equal to that entire component; if the symmetry instead swapped some stability conditions, the invariant locus would be smaller and the computed list of Donaldson-Thomas invariants might be incomplete or incorrect.","fun_headline_variants_meta":{"raw":{"variants":["Four formulas list all DT invariants on P1xP1 CY3","Complete BPS spectrum on the P1xP1 canonical bundle","Explicit DT invariants imply trivial wall-crossing on a CY3","All DT invariants on a local Calabi-Yau from quiver data"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000299,"raw_usage":{"total_tokens":1801,"prompt_tokens":1092,"completion_tokens":709,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":708,"completion_tokens_details":{"reasoning_tokens":629}},"tokens_in":708,"tokens_out":709,"duration_ms":14686,"temperature":1.0,"reasoning_tokens":629,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T17:46:06.397344+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct check would be to compute the induced $\\mathbb{Z}_2$ action on the covering $\\varpi:\\mathrm{Stab}_0(Y)\\to\\mathrm{Hom}_{\\mathbb{Z}}(\\Gamma,\\mathbb{C})$ and on the Hom-spaces among the line bundles $\\mathcal{O}_C(n)$ on the resolved conifold; if any deck transformation is nontrivial, or if the action sends the standard heart to a different heart, then the fixed locus is a proper subset of $\\mathrm{Stab}_0(Y)$ and the identification used for Theorem 4.3 fails.","supporting_citations":[{"cited_title":"Toda, Stability conditions and crepant small resolutions , Trans","cited_arxiv_id":null,"evidence_quote":"Provides the complete description of the standard connected component of the resolved conifold stability space, including its stable objects and covering of central-charge space; the invariant locus is identified with this component."},{"cited_title":"Bridgeland, A","cited_arxiv_id":null,"evidence_quote":"Gives the derived equivalence $D_c(X)\\cong D_c^H(Y)$ for the crepant resolution of a quotient, which transfers the $G$-action and the stability conditions to the local threefold."},{"cited_title":"Bryan, C","cited_arxiv_id":null,"evidence_quote":"Computes the orbifold ideal-sheaf Donaldson-Thomas invariants for zero-dimensional equivariant sheaves on the resolved conifold orbifold, giving the $-2$ and $-4$ values on the central ray."},{"cited_title":"Kontsevich and Y","cited_arxiv_id":null,"evidence_quote":"Supplies the growth-rate criterion used to conclude, from the explicitly bounded spectrum of Theorem 4.3, that the DT theory on the connected component defines an analytic wall-crossing structure."}],"review_version":1}