Nonexistence of singly compactly generated t t -structures for schemes
Anirban Bhaduri, Timothy De Deyn, Michal Hrbek, Pat Lank, Kabeer Manali-Rahul · 2025
Read Nonexistence of singly compactly generated t t -structures for schemes by Anirban Bhaduri, Timothy De Deyn, Michal Hrbek, Pat Lank, Kabeer Manali-Rahul (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-2511-01622.txt.
Introduction
Anirban Bhaduri, Timothy De Deyn, Michal Hrbek, Pat Lank, and Kabeer Manali-Rahul report a negative result in homological algebra. That field studies derived categories. These package a geometric space's sheaves into a single algebraic object. Inside such a category, a t-structure is a way of splitting objects into positive and negative parts, much like truncating a sequence.
A t-structure is singly compactly generated when one compact object controls the whole splitting, a strong finiteness property often assumed or hoped for. The authors exhibit the first schemes, geometric spaces in algebraic geometry, whose standard t-structure on the derived category of quasi-coherent sheaves lacks this property. The counterexamples close off a plausible general expectation and mark where the theory's finiteness ends.
Catalog record
| CREATOR | Anirban Bhaduri, Timothy De Deyn, Michal Hrbek, Pat Lank, Kabeer Manali-Rahul |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 26 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2511-01622
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Nonexistence of singly compactly generated t t -structures for schemes Anirban Bhaduri, Timothy De Deyn, Michal Hrbek, Pat Lank, Kabeer Manali-Rahul Abstract . We show the first instances of schemes whose standard aisles on their derived category of quasi-coherent sheaves are not singly compactly generated. 2020 Mathematics Subject Classification 14A30 (primary), 14D23, 14F08, 18E10 1. Introduction To date, the territories of derived categories associated with schemes remain largely unexplored. Researchers in these lands now possess a variety of tools at their disposal. Among these is a notion introduced by [ BBDG18 ] called a t -structure . Roughly speaking, a t -structure on a triangulated category \mathcal{T} is a pair of subcategories (\mathcal{T}^{\leq 0},\mathcal{T}^{\geq 0}) satisfying axioms that describe objects of \mathcal{T} in terms of those in \mathcal{T}^{\leq 0} and \mathcal{T}^{\geq 0} . In fact, the t -structure is completely determined by its aisle \mathcal{T}^{\leq 0} [ KV88 ] . These tools provide a coarser, and hence more digestible, topography of these lands. It remains unknown whether the standard t -structure is singly compactly generated, i.e. generated by a single compact generator, for any Noetherian scheme. Thus far, by [ Nee24 , Theorem 3.2] , all that is known is that the standard aisle D^{\leq 0}_{\operatorname{qc}}(X) is always ‘equivalent’ (in the sense of [ Nee25 , Definition 0.18] ) to a singly compactly generated aisle. To some, there was an expectation that the standard aisle was itself singly compactly generated. However, it turns out, this is horridly wrong : Theorem (see Theorem 3.8 ) . This is false for \mathbb{P}^{1}_{k} over a field k . In fact, it fails more generally for proper tame Deligne–Mumford stacks \mathcal{X} of positive Krull dimension over k . To find counterexamples is not an obvious, nor straightforward, task. Particularly, for the case of \mathbb{P}^{1}_{k} where k=\mathbb{C} , we checked the failure by using stability of vector bundles (see Appendix A ). However, the main proof in the text does not require stability, and holds far more generally. Ultimately, we show that D^{\leq 0}_{\operatorname{qc}}(\mathcal{X}) being singly compactly generated implies every object of \operatorname{coh}(\mathcal{X}) is Artinian (see Corollary 3.6 ). In fact, singly compactly generated is equivalent to \operatorname{coh}(\mathcal{X}) admitting a weak generator (see Lemma 3.4 ). And for the case of proper algebraic spaces over a field, every coherent sheaf being Artinian occurs precisely when the algebraic space is an affine Artinian scheme (see Proposition 3.7 ). Lastly, we show Noetherian schemes (more generally algebraic spaces) which are proper and of positive relative dimension over a base cannot have singly compactly generated standard aisles (see Proposition 3.11 ). For example, this includes proper schemes over the integers with positive dimensional fibers, which gives ‘arithmetic’ cases. Acknowledgment . De Deyn was supported by ERC Consolidator Grant 101001227 (MMiMMa). Hrbek was supported by the GAČR project 23-05148S and the Czech Academy of Sciences (RVO 67985840). Lank was supported under the ERC Advanced Grant 101095900-TriCatApp. Manali-Rahul is grateful to Max Planck Institute for Mathematics in Bonn for its hospitality and financial support. The authors thank Leovigildo Alonso Tarrío, Uttaran Dutta, Ana Jeremías López, Amnon Neeman, and Fei Peng for discussions and comments. 2. Preliminaries 2.1. Abelian categories Let \mathcal{A} be an essentially small abelian category. Recall that an object E\in\mathcal{A} is called Artinian (resp. Noetherian ) if any descending (resp. ascending) chain of subobjects of E becomes stationary. We say \mathcal{A} is Artinian (resp. Noetherian ) if every object is such. In the case \mathcal{A} is both Artinian and Noetherian, we say it is a length category . When \mathcal{A} is additionally Grothendieck, it is called locally Noetherian when every object of \mathcal{A} is a directed union of its Noetherian subobjects, which we denote by \operatorname{noeth}(\mathcal{A}) . Furthermore, recall an object G\in\mathcal{A} is called a weak generator if \operatorname{Hom}(G,-)\colon\mathcal{A}\to\operatorname{Ab} is faithful on objects (i.e. \operatorname{Hom}(G,E)\cong 0 implies E\cong 0 ). Lastly, if R is a commutative Noetherian ring, we say \mathcal{A} is \operatorname{Hom} -finite over R if it is an R -linear category and \operatorname{Hom}(A,B) is a finite length R -module for all A,B\in\mathcal{A} . This notion of \operatorname{Hom} -finiteness may differ from others in the literature; e.g. finitely generated module as opposed to finite length. We sometimes abusively say ‘ \operatorname{Hom} -finite’ leaving the commutative ring R implicit. 2.2. t -structures Let \mathcal{T} be a compactly generated triangulated category. We discuss t -structures on triangulated categories. See [ KV88 , BBDG18 ] for details. Denote its subcategory of compact objects by \mathcal{T}^{c} . A pair of strictly full subcategories (\mathcal{T}^{\leq 0},\mathcal{T}^{\geq 0}) of \mathcal{T} is a t -structure if \operatorname{Hom}(A,B)=0 for all A\in\mathcal{T}^{\leq 0} and B\in\mathcal{T}^{\geq 0}[-1] ; \mathcal{T}^{\leq 0}[1]\subseteq\mathcal{T}^{\leq 0} and \mathcal{T}^{\geq 0}[-1]\subseteq\mathcal{T}^{\geq 0} ; and for every E\in\mathcal{T} , there is a distinguished triangle \tau^{\leq 0}E\to E\to\tau^{\geq 1}E\to(\tau^{\leq 0}E)[1] with \tau^{\leq 0}E\in\mathcal{T}^{\leq 0} and \tau^{\geq 1}E\in\mathcal{T}^{\geq 0}[-1] . Also, we say a strictly full subcategory \mathcal{U}\subseteq\mathcal{T} is an aisle if the inclusion \mathcal{U}\to\mathcal{T} admits a right adjoint while \mathcal{U} is closed under positive shifts and extensions. In fact, for any t -structure (\mathcal{T}^{\leq 0},\mathcal{T}^{\geq 0}) , \mathcal{U} is an aisle. Particularly, any aisle \mathcal{U} determines a t -structure (\mathcal{U},\mathcal{U}^{\perp}[1]) where \mathcal{U}^{\perp}:=\{T\in\mathcal{T}\mid\forall U\in\mathcal{U},\operatorname{Hom}(U,T)=0\}. As an example, for an abelian category \mathcal{A} , we have the ‘standard’ t -structure whose aisle is given by D^{\leq 0}(\mathcal{A}) given by objects whose i -th cohomology is zero for i>0 . Given a set \mathcal{S}\subseteq\mathcal{T} , \overline{\langle\mathcal{S}\rangle}^{(-\infty,0]} is defined to be the smallest cocomplete aisle containing \mathcal{S} . By [ Nee21 , Theorem 2.3] , such aisles always exist. An aisle \mathcal{U} on \mathcal{T} is compactly generated when there exists a collection of compact objects \mathcal{P}\subseteq\mathcal{T}^{c} satisfying \overline{\langle\mathcal{P}\rangle}^{(-\infty,0]}=\mathcal{U} . Hence, we say a t -structure is compactly generated if its aisle is such. 2.3. Algebraic stacks Our conventions for algebraic stacks is [ Sta25 ] . However, we follow [ HR17 , §1] for the derived pullback/pushforward adjunction (loc. cit. follows [ Ols07 , LO08a , LO08b ] ). Also, we omit the ‘qc’ in the notation of the derived functors compared to [ HR17 ] . Symbols X , Y , etc. refer to schemes/algebraic spaces, whereas \mathcal{X} , \mathcal{Y} , etc. refer to algebraic stacks. Let \mathcal{X} be a Noetherian algebraic stack. Denote by \operatorname{Mod}(\mathcal{X}) the Grothendieck abelian category of sheaves of \mathcal{O}_{\mathcal{X}} -modules on the lisse-étale site of \mathcal{X} and \operatorname{Qcoh}(\mathcal{X}) (resp. \operatorname{coh}(\mathcal{X}) ) for the strictly full subcategory of \operatorname{Mod}(\mathcal{X}) consisting of quasi-coherent (resp. coherent) sheaves. Define D(\mathcal{X}):=D(\operatorname{Mod}(\mathcal{X})) for the (unbounded) derived category of \operatorname{Mod}(\mathcal{X}) . Set D_{\operatorname{qc}}(\mathcal{X}) (resp. D^{b}_{\operatorname{coh}}(\mathcal{X}) ) for the full subcategory of D(\mathcal{X}) consisting of complexes with quasi-coherent cohomology sheaves (resp. which are bounded and have coherent cohomology sheaves). Moreover, \operatorname{Perf}(\mathcal{X}) is the full subcategory of perfect complexes in D_{\operatorname{qc}}(\mathcal{X}) which can be defined for any ringed site [ Sta25 , Tag 08G4 ] , and so, in particular, for \mathcal{X} by looking at its lisse-étale site. As a warning, the compact objects of D_{\operatorname{qc}}(\mathcal{X}) are perfect complexes [ HR17 , Lemma 4.4] , although the converse need not be true. We say \mathcal{X} is affine-pointed if every morphism \operatorname{Spec}(k)\to\mathcal{X} from a field k is affine; this is automatic when \mathcal{X} has quasi-affine or quasi-finite diagonal. Furthermore, \mathcal{X} is said to satisfy