A configuration space model for algebraic function spaces To Benson Farb on his 57 th birthday
Oishee Banerjee · 2025
Read A configuration space model for algebraic function spaces To Benson Farb on his 57 th birthday by Oishee Banerjee (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-00105.txt.
Introduction
Banerjee proves a result in algebraic geometry, published in 2024, about spaces of maps between smooth shapes defined by equations. The paper shows that, under stated conditions, such a space of algebraic maps admits a model built from configurations of points. The result carries a known theorem about ordinary topological map spaces into the algebraic setting, and extends the author's earlier work to higher dimensions.
Catalog record
| CREATOR | Oishee Banerjee |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 49 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-00105
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A configuration space model for algebraic function spaces To Benson Farb on his 57 th birthday Oishee Banerjee Abstract We prove that the space of algebraic maps between two smooth projective varieties, under certain conditions, admit a configuration space model, thereby obtaining an algebro-geometric analogue of Bendersky-Gitler’s result ( [ BG91 , Theorem 7.1] ) on topological function spaces. Our result should be a thought of as a natural higher dimensional counterpart of [ Ban24 , Theorem 3] . 1 Introduction Motivation. The study of spaces of continuous maps between two topological spaces, and their surprising connections with configuration spaces, has a rich history spanning several decades. Foundational works by Anderson, Bendersky-Gitler, Snaith, Cohen-May-Taylor, Arone, Ahearn-Kuhn, and others have explored this extensively (see [ BG91 , And72 ] , and also [ AK02 , Aro99 ] and the references therein). One key result is the stable splitting of function spaces under certain constraints (e.g., connectivity conditions on the range), where the components of the splitting include, among other structures, configuration spaces on the domain. An exact analogue of this phenomenon for algebraic maps between algebraic varieties is unrealistic due to the rigidity of morphisms between varieties. Nevertheless, in this note, we show that the moduli space of algebraic morphisms between two smooth projective varieties, under certain strong conditions on the range, can, in a sense, admit a configuration space model. This establishes an algebro-geometric analogue of Bendersky-Gitler’s result on the space of continuous maps between topological spaces ( [ BG91 , Theorem 7.1] ). Our result can be seen as a natural higher-dimensional counterpart of [ Ban24 , Theorem 3] . Setup. Throughout the paper, we fix smooth projective varieties X and Y over an algebraically closed field of characteristic 0 . We fix \Upsilon a polarization on Y i.e. we fix an embedding \upsilon:Y\hookrightarrow\mathbb{P}^{N} where \Upsilon=\upsilon^{*}\mathcal{O}_{\mathbb{P}^{N}}(1) and N=\dim|\Upsilon| , the rank of the complete linear system |\Upsilon| . A morphism f:X\to\mathbb{P}^{N} corresponds to a line bundle L on X such that L=f^{*}\mathcal{O}_{\mathbb{P}^{N}}(1) . Let \mathbf{d}:=c_{1}(L)\in N^{1}(X) . A morphism f:X\to Y corresponds to a line bundle L on X such that L=f^{*}\Upsilon={f^{*}}{\circ}\upsilon^{*}\mathcal{O}_{\mathbb{P}^{N}}(1) . We say f has degree \mathbf{d} if c_{1}(L)=\mathbf{d} . Let \mathrm{Mor}_{\mathbf{d}}(X,Y) be the moduli space of morphisms f:X\to Y such that c_{1}(f^{*}\Upsilon)=\mathbf{d} . We say a numerical class \mathbf{d}\in N^{1}(X) separates r points if it is ample and if every line bundle L\in\mathrm{Pic}_{\mathbf{d}}(X) separates r points 1 1 1 There are closely related (stronger) notions like r -very ampleness, r -spannedness etc. which have been studied for decades, pioneered by the work of Beltrametti, Sommese and others, however, the weaker notion of separating points (as opposed to jets) is sufficient for the purpose of this note. Aumonier in [ Aum24 ] calls this property of line bundles k -interpolating. However, ‘separating points’ is a standard phrase in algebraic geometry to describe the phenomenon, so we adhere to that. (see 2.3.1 ). Note that if a line bundle L (respectively, a numerical class \mathbf{d} ) separates r points, then it separates (r-1) points; let r(L) (respectively, r(\mathbf{d}) ) denote the maximum number r for which L (respectively, \mathbf{d} ) separates r -points. If \mathbf{d}\in N^{1}(X) separates at least one point, we define r(\mathbf{d}):=\max\{r:\mathbf{d}\text{ separates }r\text{ points}\}. (1.0.1) Furthermore, we say \mathbf{d} is acyclic of all higher cohomologies of all line bundles in its numerical equivalence class vanish. Now we state our theorem. Theorem 1.0.1 . Let X be a smooth projective variety of dimension n , (Y,\Upsilon) a polarized smooth projective variety, and N:=\dim|\Upsilon| . Let \mathbf{d} be acyclic, and let r(\mathbf{d}) be as in ( 1.0.1 ). Then: 1. Then there exists a first quadrant spectral sequence of Galois representations/mixed Hodge structures: \displaystyle E_{1}^{p,q}\implies H^{p+q}_{c}(\mathrm{Mor}_{\mathbf{d}}(X,Y);\mathbf{Q}) (1.0.2) with \displaystyle E_{1}^{p,*}=(H^{*}(\mathring{X}^{p};\mathbf{Q})\otimes\mathit{sgn}_{S_{p}})^{S_{p}}\otimes H^{*}(\mathrm{Pic}_{\mathbf{d}}(X);\mathbf{Q})\otimes H_{c}^{*}(Y(D_{p-1});\mathbf{Q}) (1.0.3) for all p\leq r(\mathbf{d})+1 , where Y(D_{p}) are certain auxilliary schemes defined in, and satisfying the conditions of § 2.5.2 , and \mathring{X}^{p}:=X^{p}-\{\text{ diagonals}\}. 2. Homological stability: In the case when Y=\mathbb{P}^{N} , the spectral sequence ( 1.0.2 ) results in \displaystyle E_{2}^{p,q}=E_{\infty}^{p,q} (1.0.4) for all 0\leq p\leq r(\mathbf{d})+1 , and \dim(\mathrm{Mor}_{\mathbf{d}}(X,Y))-r(\mathbf{d})\leq q\leq\dim(\mathrm{Mor}_{\mathbf{d}}(X,Y)) . 3. If \delta:=\mathbf{d}-c_{1}(K_{X}) is ample, then the stable bound r(\mathbf{d}) satisfies the following equality: r(\mathbf{d})=\left\lfloor\min_{\begin{subarray}{c}[W]\in\mathrm{CH}_{k}(X),\\ 1\leq k\leq n\end{subarray}}\frac{2(\delta^{k}.[W])^{\frac{1}{k}}-n^{2}+n-1}{2}\right\rfloor-1 (1.0.5) where \mathrm{CH}_{k}(X) is the k^{th} Chow group of X . Some remarks and context 1. Poincare/Koszul duality. A recurring phenomenon in the study of such algebraic function spaces, as demonstrated in this note, as well as in [ Ban24 ] , is the appearance of a Koszul-type cochain complex (see ( 4.1.5 )) in the analysis of the Poincare dual of \mathrm{Mor}_{\mathbf{d}}(X,Y) . Whereas it appears naturally in our proof via the theory of hypercovers, it does beg the question of whether there is a factorization homology approach to proving something like Theorem 1.0.1 . An affirmative answer has already been provided by Ho ( [ Ho20 ] ) in the case X=\mathbb{P}^{1} and Y=\mathbb{P}^{n} (because in such cases, algebraic maps simply boil down to studying zero-cycles on X ) and there is paramount evidence of an affirmative answer for curves of higher genera by exploiting algebraic non-abelian Poincare duality in the sense of Gaitsgory-Lurie ( [ GL17 , §3] ). However, a naive translation of the curve-case to higher dimensional domains have several obvious pitfalls, some of which becomes clear as we go through the proof, and discussing the rest would take us too far afield (see e.g. [ GL19 , §1.5] ). 2. \mathrm{Mor}_{\mathbf{d}}(X,Y) vs. its topological analogue \mathrm{Top}_{\mathbf{d}}(X,Y) . In a direction distinct from the study of the relationship between (continuous) function spaces and configuration spaces, significant attention has been given to the comparison of spaces of holomorphic/algebraic functions with that of continuous functions between complex holomorphic manifolds since the ’70s (for a brief history of it see [ Aum24 , §1.1] ). Two notable examples are: Mostovoy’s work ( [ Mos06 ] ) where X,Y are both projective spaces, and a recent work of Aumonier’s (see [ Aum24 ] ) where he compares stable homology of \mathrm{Mor}_{\mathbf{d}}(X,\mathbb{P}^{N}) to that of continuous maps X\to\mathbb{P}^{N} for arbitrary smooth projective X , thus significantly extending Mostovoy’s result. If our theorem relating the (cohomology of the) space of algebraic maps to configuration spaces appears somewhat unexpected, it is worth highlighting that results by Anderson, Bendersky-Gitler ( [ And72 , BG91 ] ), when combined with those of Aumonier or Mostovoy, lend credence to such a connection. Our theorem not only confirms this connection but also strengthens it in two key ways: (a) it demonstrates that the Galois representations or mixed Hodge structures on both sides are preserved, and (b) it underscores the role of the intersection theory of X , as evidenced by Equation 1.0.5 . Notably, pulling back a ’configuration space model’ for the space of continuous maps via Segal-type results from Aumonier or Mostovoy offers no insight into Hodge structures. In contrast, our approach—grounded entirely in algebraic geometry—makes these structures explicit and central to the discussion. The only prior instance of an explicit comparison between the space of algebraic function spaces and configuration spaces appears in the author’s earlier work [ Ban24 , Ban22 ] — those were in the case when X is a curve. 3. Adjoint line