{"id":"5c8717e5-e6e5-4ee3-b8b7-72da1ef37f0a","arxiv_id":"2608.08680","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"Under Kirwan surjectivity, the period map from algebraic asymptotic hyperkähler structures on a hyperkähler quotient to a period chamber is a bijection.","lead":"This paper proves a Torelli-type theorem: for hyperkähler quotients of a quaternionic vector space, the moduli space of algebraic asymptotic hyperkähler metrics is bijectively parametrized by a chamber of period triples, assuming the Kirwan map is surjective. It generalizes Kronheimer's ALE classification to toric hyperkähler and Nakajima quiver varieties.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The surjectivity proof depends on Prop. 2.11, whose inference from vanishing of the period of a linear combination of curve classes to vanishing on a single generator is invalid; if D'≠D, Eq. (4.16) fails and the claimed Torelli bijection is unsupported.","rationale":"The reader's weakest assumption identifies the same load-bearing concern: Prop. 2.11's identification of the discriminant locus D' with the wall D is used essentially in Theorem 4.37, step (3), through Eq. (4.16), and the proof of Prop. 2.11 contains a genuine logical gap. The paper's central claim is a bijectivity statement, and the construction of a metric realizing an arbitrary period ξ∈Ω requires that the set of parameters avoiding D'_R map onto the set of periods avoiding D_R. Without the wall-discriminant identity, the preimage of ξ under the Kirwan map may lie inside the discriminant, and the quotient X_ζ may fail to be diffeomorphic to the chosen resolution Y_{β0}; hence the surjectivity half of the Torelli theorem is not established. This is not a matter of disagreement with existing consensus; it is an internal gap in the argument. I therefore concur with the CONDITIONAL verdict and do not propose changing it. A secondary concern, the extension of algebraicity from rational to irrational parameters by continuity in Lemma 4.38, is also present, but the wall-discriminant issue is the most directly load-bearing: without Eq. (4.16), the surjectivity construction does not even produce a candidate asymptotic structure for the targeted periods.","tokens_in":26336,"tokens_out":15920,"duration_ms":192335,"concrete_test":"Test Prop. 2.11 on the minimal resolution of the A_2 Du Val singularity. Fix generators E_1,E_2 of H_2(Y), and choose c∈H^2(Y;C) in the universal Poisson deformation base such that c(E_1)=1 and c(E_2)=-1, so the class E_1+E_2 has zero period while neither generator does. Determine whether the fiber Y_c is affine. Prop. 2.11 as written predicts Y_c is affine, because c lies in neither hyperplane E_1^⊥ nor E_2^⊥. If the fiber is non-affine, the proposition is false. If the fiber is affine, the proof's inference from ∫_{E_1+E_2}Ω_c=0 to a vanishing integral on a single generator is still invalid, so Eq. (4.16) remains unproven without a new argument.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Proposition 2.11 is the load-bearing step: Theorem 4.37's surjectivity proof uses it through Eq. (4.16), D'_R = κ_{β0,R}^{-1}(D_R). The proof of Prop. 2.11 contains an invalid linear-algebra implication. If c∈D', the proof produces an actual curve Σ in an exceptional fiber with ⟨Σ,Ω_c⟩=0. Lemma 2.9 only gives [Σ] = Σ c_α[Σ_α] in the span of the chosen generators of H_2(Y_0); from Σ c_α⟨Σ_α,Ω_c⟩=0 one cannot infer that ⟨Σ_α,Ω_c⟩=0 for any single generator. The converse direction has the same gap in reverse: ⟨Σ_α,Ω_c⟩=0 shows that the class [Σ_α] is of type (1,1) with respect to I_c, not that the cycle Σ_α is holomorphic in Y_c, so it does not establish that Y_c is non-affine. Consequently the asserted identity D'=D is unsupported as stated. If it fails, then for some period ξ∈Ω with ξ_I,ξ_J∉D_R, every ζ satisfying κ_{β0,R}(ζ)=ξ may have ζ_I or ζ_J in D'_R, so X_ζ is not diffeomorphic to the reference resolution Y_{β0}; the construction in Theorem 4.37, step (3), then produces no algebraic asymptotic metric with period ξ, and surjectivity of p is not proved. Thus the central Torelli bijection rests on an unproven wall-discriminant identification.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a Torelli-type theorem for the moduli space of algebraic asymptotic hyperkähler structures on hyperkähler quotients of a quaternionic vector space by a compact group G⊂Sp(n). The main result, Theorem 4.37, states that under assumptions on the stable locus, free action of GC, and surjectivity of the Kirwan map, the hyperkähler quotient Y^{β0} is a projective crepant resolution of the central quotient X0 and the period map p:M→Ω is bijective. The proof combines a principal twistor model from the author's previous work to obtain injectivity, a period-domain computation using a wall D defined by vanishing periods, and a surjectivity argument that compares the discriminant locus of a Poisson deformation with the pullback of D via the Kirwan map. The paper also claims applications to toric hyperkähler varieties, Nakajima quiver varieties, and ALE gravitational instantons.","tokens_in":26609,"tokens_out":3803,"duration_ms":39964,"significance":"If the main theorem and its proof were correct, the paper would provide a significant generalization of Kronheimer's Torelli theorem to higher-dimensional hyperkähler quotients, with concrete classes of examples. The reduction of the period-map surjectivity to Kirwan surjectivity is a natural and potentially useful strategy, and the explicit treatment of toric and quiver examples is valuable. However, the central proof contains several load-bearing gaps: the identification of the discriminant locus with the wall D (Proposition 2.11) is logically flawed, equation (4.16) is asserted rather than proved, and the proof of algebraicity of the constructed metrics by continuity (Lemma 4.38) is not valid. In addition, the injectivity part depends entirely on the author's unpublished preprint [10]. These issues mean that the paper's main claim is not established at the present level of rigor, although the overall framework is promising.","major_comments":[{"comment":"The proof of Proposition 2.11 contains an invalid linear-algebra inference. Given c∈D′, the proof produces a curve Σ in an exceptional fiber with ⟨Σ,Ω_c⟩=0. Lemma 2.9 only gives [Σ]=∑ c_α[Σ_α] in the span of the chosen generators of H_2(Y_0). From ∑ c_α⟨Σ_α,Ω_c⟩=0 one cannot conclude that ⟨Σ_α,Ω_c⟩=0 for any single generator Σ_α, so the conclusion [Ω_c]∈H_α⊂D does not follow. The converse direction has a complementary gap: from ⟨Σ_α,Ω_c⟩=0 one only learns that the class [Σ_α] is of type (1,1) with respect to I_c, not that the cycle Σ_α is holomorphic in Y_c; hence the conclusion that Y_c is non-affine is unsupported. Because Proposition 2.11 is the foundation for the identity D′=D used throughout the paper, this gap affects the definition of the period domain and the proof of surjectivity in Theorem 4.37.","section":"§2.1, Proposition 2.11"},{"comment":"The identity D′_R = κ_{β0,R}^{-1}(D_R) is asserted in Step (3) with the justification that it follows from Proposition 2.11. This is a load-bearing step: it is precisely what allows the author to lift a period ξ∈Ω with ξ_I,ξ_J∉D_R to a parameter ζ∈Ω_IJ. Since Proposition 2.11 is not established (see previous comment), equation (4.16) is not proven. Moreover, even if Proposition 2.11 were true for the universal Poisson deformation, one would need a separate argument showing that the discriminant locus of the family Y^{β} constructed in Proposition 4.25 is compatible with the pullback of the wall under the Kirwan map; this compatibility is not proved. As a result, the surjectivity of the period map p:M→Ω is not demonstrated.","section":"§4.4, Theorem 4.37, step (3), Eq. (4.16)"},{"comment":"The proof of algebraicity of the push-forward metric (Φ_ζ)_*g_ζ is not valid. Step (3) argues that because rational points are dense in Ω_IJ and the family of diffeomorphisms {Φ_ζ} varies smoothly, the twistor space for the push-forward metric must coincide with the algebraic twistor space Z^{β0}_ζ for all ζ∈Ω_IJ by continuity. Algebraicity of a twistor space is not a closed or continuous condition in any evident sense; a limit of algebraic objects over a parameter space need not be algebraic without additional structural control. The author needs a direct construction of the algebraic twistor space for all ζ, or a formal deformation-theoretic argument, rather than an appeal to density and continuity. Since algebraicity is part of the definition of the moduli space M, this gap directly affects the claim that the constructed metrics lie in M.","section":"§4.4.2, Lemma 4.38"},{"comment":"The injectivity of the period map is imported from the author's unpublished preprint [10, Theorem 3.40, Cor. 4.12]. This dependence is load-bearing: the entire uniqueness half of the Torelli theorem rests on results that are not available to the reader in the present manuscript. For a journal submission, relying on an unpublished arXiv preprint for a central theorem is a significant concern. The author should either include the necessary statements and proofs, or state clearly which parts of [10] are being assumed and indicate their status. This is not merely a matter of exposition; it affects the verifiability of the main claim.","section":"§3.1, Proposition 3.2"}],"minor_comments":[{"comment":"The notation H2(Y;C)=⟨[Σ_α]|Σ_α⊂Y_0⟩ should be H_2(Y;C), and the phrase 'algebraic curve' could be made precise (e.g., compact one-dimensional subvariety). This is a minor notational issue but affects readability.","section":"§2.1, Definition 2.10"},{"comment":"In the displayed formula for φ_β, the expression Y^β := µ_C^{-1}(z^*_C)//_{β}G_C is ambiguous; it should be the family over z^*_C whose fiber over α is µ_C^{-1}(α)//_β G_C. Clarifying this would avoid confusion.","section":"§4.2.1, Proposition 4.25"},{"comment":"The notation F_β for the smooth trivialization and Φ_{β,β0} for the diffeomorphism is introduced but not used consistently in the proof of Theorem 4.37. In particular, in (4.13) the composition is written as Φ_{β,β0}×id after F_β, but the domains and codomains are not spelled out. Please make the diagram precise.","section":"§4.3.2, Remark 4.36(2)"},{"comment":"The proof that the stabilizer is trivial assumes that the set of column vectors corresponding to nonzero components spans R^d and then chooses d vectors forming a Z-basis. This uses unimodularity, but the link between 'span over R' and 'span over Z' should be stated more explicitly, since not every set spanning R^d contains a Z-basis of Z^d. The argument appears fixable, but as written it is too quick.","section":"§5.1, Lemma 5.4"}],"recommendation":"major_revision","confidential_remarks":"The paper addresses a worthwhile question and the proposed reduction to Kirwan surjectivity is attractive. However, the referee report identifies two central mathematical gaps (Proposition 2.11 and Lemma 4.38) and an unproved identity (4.16) that underpins surjectivity. These are not merely cosmetic issues; they concern the logical validity of the main theorem. The dependence on the unpublished preprint [10] for injectivity is also a concern for a journal submission. I recommend major revision rather than rejection because the framework and examples are promising, and it is conceivable that the gaps can be repaired within the scope of the paper. The author should, at minimum, prove Proposition 2.11 under additional hypotheses or replace it with a different wall-discriminant identification, give a rigorous algebraic proof of Lemma 4.38, and clarify the status of the results from [10]."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. First, the paper is a synthesis: injectivity of the period map is imported from the author's PTM preprint [10], and the new content is the surjectivity reduction for hyperkähler quotients via the Kirwan map and the chamber structure. Second, the main theorem is not proven as written. The load-bearing step, Proposition 2.11, identifies the discriminant D' of the universal Poisson deformation with the wall D defined by vanishing periods of generating curve classes. The proof does not work: from vanishing of the period of a linear combination of curve classes you cannot conclude any single generator has vanishing period, and the converse direction confuses a class being of type (1,1) with the cycle being holomorphic in the deformed complex structure. If D' ≠ D, then Eq. (4.16) fails and the surjectivity argument in Theorem 4.37 has no bridge from periods on the quotient side to the period domain on the resolution.\n\nThe paper does real work elsewhere. The framework is clearly laid out: conical symplectic varieties, Poisson deformations, the period domain, and the role of each assumption is stated (Remark 4.6). The examples section is concrete and useful. For coloop-free unimodular toric data and strict Schur root quiver data, the stable-locus and free-action conditions are proved, and the Kirwan surjectivity is cited to Konno and McGerty–Nevins. So the conditional statement is well motivated and, if the gaps are repaired, would genuinely generalize Kronheimer's ALE Torelli theorem to toric and quiver varieties.\n\nThe second soft spot is Lemma 4.38, where algebraicity of the push-forward metric is extended from rational parameters to all real parameters by continuity. That is not a valid argument: the twistor space is not known to vary continuously in the required sense, and algebraicity is not a closed condition. The rational-point case is also sketched rather than proved. This matters because the theorem asserts the constructed metrics are algebraic.\n\nOne more dependency worth flagging: injectivity comes from the unpublished preprint [10]. That is not a flaw by itself, but it means the paper is not self-contained in a central way, and a referee needs [10] in hand.\n\nWho gets value from this? Specialists in asymptotic hyperkähler metrics and symplectic resolutions will find the framework and the example computations worth reading. The paper deserves a serious referee: the conditional statement is significant and the flaws are identifiable, not hopeless. I would send it to review with instructions to focus on Prop 2.11 and Lemma 4.38, and to ask the author to either prove the wall-discriminant identification or make it an explicit assumption. My own verdict is that the theorem is unproven as it stands, but the work is worth engaging.","headline":"A conditional Torelli theorem for hyperkähler quotients that is plausible but not yet proven: the surjectivity step rests on an unproved identification between the discriminant of the universal Poisson deformation and the wall defined by curve periods.","tokens_in":27216,"tokens_out":6492,"would_cite":false,"duration_ms":68181,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14J42","53C26","14L24","14D21","14E15"],"pacs":[],"model":"deepseek-v4-flash","headline":"Assuming a surjective Kirwan map, the period map is a bijection for algebraic asymptotic hyperkähler structures on hyperkähler quotients.","keywords":["Torelli-type theorem","period map","hyperkähler quotient","Kirwan map","crepant resolution","Poisson deformation","Nakajima quiver variety","toric hyperkähler variety"],"falsifier":"Look for a hyperkähler quotient satisfying the paper's standing assumptions for which the discriminant locus of the universal Poisson deformation is not equal to the wall $D$ formed by vanishing periods of the individual generating curves; if such an example exists, the identity $D'_{\\mathbb{R}}=\\kappa^{-1}_{\\beta_0,\\mathbb{R}}(D_{\\mathbb{R}})$ used in Step (3) of Theorem 4.37 fails and the surjectivity proof does not go through.","tokens_in":26046,"feed_emoji":"🔁","tokens_out":8885,"duration_ms":84051,"temperature":0.7,"pith_summary":"This paper is trying to prove a Torelli-type theorem: for hyperkähler quotients of quaternionic vector spaces $\\mathbb{H}^n$ by compact subgroups of $\\mathrm{Sp}(n)$, the moduli space of algebraic hyperkähler structures asymptotic to the central cone metric is classified completely by the triple of Kähler-class periods. Provided the stable locus is non-empty, a generic lattice point has a free complexified action, and the Kirwan map is surjective, the period map is a bijection onto the period domain. If correct, the result would generalize the classical Torelli theorem for ALE gravitational instantons and would apply uniformly to toric hyperkähler varieties and Nakajima quiver varieties. The paper's contribution is to make surjectivity algebro-geometric: it identifies the periods of the quotient metrics with the Kirwan map on rational parameters and then extends to the whole chamber by continuity.","feed_headline":"Period map is bijective for hyperkähler quotients","feed_subtitle":"One surjectivity assumption makes each triple of Kähler classes realized by exactly one asymptotic hyperkähler structure.","key_machinery":"The principal twistor model is the fibration over $\\mathbb{C}\\otimes\\mathcal{O}(2)$ built from the universal Poisson deformation of the crepant resolution, and its universality theorem provides the injectivity of the period map. The wall $D$ is the union of hyperplanes $H_\\alpha=\\{\\langle \\Sigma_\\alpha,\\Omega_c\\rangle=0\\}$ indexed by algebraic curves $\\Sigma_\\alpha$ generating $H_2(Y_0)$; the period domain $\\Omega$ is the complement of $D_{\\mathbb{R}}\\otimes\\mathbb{R}^3$ in $H^2(Y;\\mathbb{R})\\otimes\\mathbb{R}^3$. The Kirwan map $\\kappa_\\beta:z^*_{\\mathbb{C}}\\to H^2(Y^\\beta;\\mathbb{C})$ records Chern classes of the line bundles associated to characters, and the load-bearing identity is $\\hat p=\\kappa_{\\beta_0,\\mathbb{R}}\\otimes\\mathbb{R}^3$ on the dense rational set $\\Omega^{\\mathbb{Q}}_{IJ}$, extended by continuity, together with the equality $D'_{\\mathbb{R}}=\\kappa^{-1}_{\\beta_0,\\mathbb{R}}(D_{\\mathbb{R}})$ supplied by Proposition 2.11. This identity converts the algebraic classification of Poisson-deformation periods into the analytic construction of quotient metrics.","core_discovery":"The paper claims Theorem 4.37: for a hyperkähler quotient of $\\mathbb{H}^n$ by a compact subgroup $G\\subset \\mathrm{Sp}(n)$, if the stable locus of the complex moment map at zero is non-empty and there is a generic lattice point $\\beta_0$ where $G_{\\mathbb{C}}$ acts freely and the Kirwan map $\\kappa_{\\beta_0}$ is surjective, then $Y^{\\beta_0}$ is a projective crepant resolution of the central quotient $X_0$ and the period map $p:\\mathcal{M}\\to\\Omega$ is bijective. The proof splits into injectivity, obtained from the universality of the principal twistor model; image containment, obtained from metric non-degeneracy together with the wall definition; and surjectivity, which is the hard part. On rational moment-map parameters the period map equals $\\kappa_{\\beta_0,\\mathbb{R}}\\otimes\\mathbb{R}^3$, this equality extends to the real chamber by continuity, and the discriminant locus of the universal Poisson deformation coincides with the inverse image of the wall under the Kirwan map, so every period triple in $\\Omega$ is realized by some hyperkähler quotient metric pushed forward to $Y^{\\beta_0}$ and shown algebraic by the twistor-space argument.","pith_inferences":["The proof suggests a testable extension beyond the paper's hypotheses: if the Kirwan map is not surjective, the period image should be exactly the intersection of $\\Omega$ with the real span of the Kirwan image, rather than all of $\\Omega$.","The wall-coincidence assertion in Proposition 2.11 is the fragile step; computing the discriminant locus of the universal Poisson deformation for toric hyperkähler varieties with non-generic weight matrices would test whether the period domain needs to be redefined using the actual discriminant locus instead of the vanishing-period wall.","For conical symplectic varieties that are not hyperkähler quotients, the same continuity argument could show that $p(\\mathcal{M})$ is a non-empty open and closed subset of $\\Omega$, which would give surjectivity without any Kirwan map.","The twistor-space algebraicity proof indicates that the metric realization varies continuously with the moment-map parameter, so small perturbations of a period triple should produce nearby asymptotic hyperkähler metrics; this could be checked concretely in the toric case."],"forward_implications":["Every period triple in the chamber $\\Omega$ is realized by exactly one isomorphism class of algebraic asymptotic hyperkähler structures on the projective crepant resolution $Y^{\\beta_0}$.","The theorem yields Torelli-type classifications for toric hyperkähler varieties attached to coloop-free unimodular matrices and for Nakajima quiver varieties whose dimension vectors form a strict Schur root.","It provides an algebro-geometric proof of the classical Torelli theorem for ALE gravitational instantons, covering the analytic classification by a different route.","Under the same assumptions, the twistor space $Z_s$ associated to a real section $s$ is a twistor space if and only if $s$ is not contained in the wall $D$, answering a question left open in the companion paper on principal twistor models.","Any hyperkähler quotient satisfying the wall equality and Kirwan surjectivity automatically has a bijective period map, so the mechanism is reusable beyond the specific examples worked out here."],"supporting_citations":[{"why":"Introduces the principal twistor model and its universality theorem; the paper relies on it for the injectivity of the period map and the moduli setup.","marker":"[10]"},{"why":"Provides the Kempf–Ness type identification of hyperkähler quotients with GIT quotients, used to view $Y^\\beta$ as a projective crepant resolution.","marker":"[7]"},{"why":"Shows the complex moment map defines a Poisson deformation whose period map equals the Kirwan map; this is the surjectivity input.","marker":"[15]"},{"why":"Establishes the universal Poisson deformation and the period-map isomorphism identifying its base with $H^2(Y;\\mathbb{C})$, grounding the wall and period-domain framework.","marker":"[17]"},{"why":"Proves the pure Hodge structure on fibers of a crepant resolution, used to generate $H^2$ by algebraic curves and to define the wall $D$.","marker":"[6]"},{"why":"Supplies the analytic crepant-resolution statement used to control non-projective hyperkähler quotients and their asymptotic metrics.","marker":"[16]"},{"why":"The hyperkähler quotient construction of ALE spaces whose Torelli theorem is the classical case being generalized.","marker":"[11]"},{"why":"Confirms the surjectivity of the Kirwan map for toric hyperkähler varieties, making the main theorem applicable there.","marker":"[9]"},{"why":"Confirms the Kirwan surjectivity for Nakajima quiver varieties, making the main theorem applicable there.","marker":"[14]"}],"fun_headline_variants":["Torelli for hyperkähler quotients","Bijective period map on hyperkähler quotients","One condition enough for Torelli on hyperkähler quotients","Hyperkähler quotients: Torelli theorem","Period map bijective on hyperkähler quotients"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof assumes that if the holomorphic symplectic form on a fiber pairs to zero against a linear combination of the generating curve classes, then it must already pair to zero against one of the individual generators; the surjectivity of the period map depends on this step, and the implication is not generally valid.","fun_headline_variants_meta":{"raw":{"variants":["Torelli for hyperkähler quotients","Bijective period map on hyperkähler quotients","One condition enough for Torelli on hyperkähler quotients","Hyperkähler quotients: Torelli theorem","Period map bijective on hyperkähler quotients"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000792,"raw_usage":{"total_tokens":3480,"prompt_tokens":924,"completion_tokens":2556,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":540,"completion_tokens_details":{"reasoning_tokens":2473}},"tokens_in":540,"tokens_out":2556,"duration_ms":22506,"temperature":1.0,"reasoning_tokens":2473,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T04:28:39.198044+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Look for a hyperkähler quotient satisfying the paper's standing assumptions for which the discriminant locus of the universal Poisson deformation is not equal to the wall $D$ formed by vanishing periods of the individual generating curves; if such an example exists, the identity $D'_{\\mathbb{R}}=\\kappa^{-1}_{\\beta_0,\\mathbb{R}}(D_{\\mathbb{R}})$ used in Step (3) of Theorem 4.37 fails and the surjectivity proof does not go through.","supporting_citations":[{"cited_title":"Principal twistor models and asymptotic hyperk\\\"ahler metrics","cited_arxiv_id":"2603.03923","evidence_quote":"Introduces the principal twistor model and its universality theorem; the paper relies on it for the injectivity of the period map and the moduli setup."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the Kempf–Ness type identification of hyperkähler quotients with GIT quotients, used to view $Y^\\beta$ as a projective crepant resolution."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Shows the complex moment map defines a Poisson deformation whose period map equals the Kirwan map; this is the surjectivity input."},{"cited_title":"Namikawa","cited_arxiv_id":null,"evidence_quote":"Establishes the universal Poisson deformation and the period-map isomorphism identifying its base with $H^2(Y;\\mathbb{C})$, grounding the wall and period-domain framework."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Proves the pure Hodge structure on fibers of a crepant resolution, used to generate $H^2$ by algebraic curves and to define the wall $D$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"The hyperkähler quotient construction of ALE spaces whose Torelli theorem is the classical case being generalized."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Confirms the surjectivity of the Kirwan map for toric hyperkähler varieties, making the main theorem applicable there."},{"cited_title":"McGerty and T","cited_arxiv_id":null,"evidence_quote":"Confirms the Kirwan surjectivity for Nakajima quiver varieties, making the main theorem applicable there."}],"review_version":1}