Classifying thick subcategories over a Koszul complex via the curved BGG correspondence

Jian Liu, Josh Pollitz · 2025

Read Classifying thick subcategories over a Koszul complex via the curved BGG correspondence by Jian Liu, Josh Pollitz (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-2502-13806.txt.

Introduction

Jian Liu and Josh Pollitz classify certain algebraic subcategories in this 2025 mathematics paper. The setting is a Koszul complex built from a list of elements in a regular ring. The objects classified are the thick subcategories of a derived category of differential graded modules. A curved version of a classic correspondence carries the proof.

Their result recovers a theorem of Stevenson when the element list is a regular sequence. It also covers the known classification for an exterior algebra, through that same correspondence. A key step classifies tensor submodules over graded rings, recovering a result of Hopkins and Neeman. One payoff: the lattice of these subcategories stays fixed under Grothendieck duality. The paper unifies several older classification theorems under one frame.

Catalog record

CREATORJian Liu, Josh Pollitz
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE104 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-13806

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Classifying thick subcategories over a Koszul complex via the curved BGG correspondence Jian Liu, Josh Pollitz Abstract . In this work we classify the thick subcategories of the bounded derived category of dg modules over a Koszul complex on any list of elements in a regular ring. This simultaneously recovers a theorem of Stevenson when the list of elements is a regular sequence and the classification of thick subcategories for an exterior algebra over a field (via the BGG correspondence). One of the major ingredients is a classification of thick tensor submodules of perfect curved dg modules over a graded commutative noetherian ring concentrated in even degrees, recovering a theorem of Hopkins and Neeman. We give several consequences of the classification result over a Koszul complex, one being that the lattice of thick subcategories of the bounded derived category is fixed by Grothendieck duality. 2020 Mathematics Subject Classification 13D09 (primary); 14M10, 18G80 (secondary) Introduction The problem of classifying thick (or localizing) subcategories of a given triangulated category has maintained a great deal of interest—in a variety of different areas—since the work of Devinatz, Hopkins, and Smith [ 22 ] in stable homotopy theory; see, for instance, [ 11 , 12 , 21 , 27 , 33 , 40 , 41 , 43 , 45 ] . Perhaps the most relevant to the present article is a theorem of Hopkins [ 27 ] and Neeman [ 33 ] that establishes a one-to-one correspondence between the thick subcategories of the homotopy category of perfect complexes over a commutative noetherian ring and the specialization closed subsets of the prime spectrum of the ring. Thomason [ 45 ] later extended this result to quasi-compact and quasi-separated schemes, provided the thick subcategories are closed under a certain tensor action. These foundational results have been widely applied in both algebra and topology. Let R be a commutative noetherian ring, and write \Df(R) for its bounded derived category; the objects are R -complexes with finitely generated total homology and morphisms are obtained by formally inverting quasi-isomorphisms. The theorem of Hopkins–Neeman yields a classification of the thick subcategories of \Df(R) when R is a regular ring. In the singular setting, the most notable classification of the thick subcategories of \Df(R) is due to Stevenson [ 40 ] which establishes a classification when R is a complete intersection in the sense that it is a regular ring modulo a regular sequence; see also [ 16 ] . In this article we are concerned with a more general setting that recovers Stevenson’s result as well as the classification of thick subcategories over an exterior algebra over a field simultaneously. Let E be a Koszul complex on a list of elements f_{1},\ldots,f_{n} regarded as a dg R -algebra in the usual way; cf. Section 4 . Let \Df(E) denote the derived category of dg E -modules whose homology is finitely generated over R . The main result of the article classifies the thick subcategories of \Df(E), when R is a regular ring; specializing to when {\bm{f}} is an R -regular sequence recovers Stevenson’s theorem, and specializing to {\bm{f}}=0 one obtains the classification of thick subcategories for an exterior algebra over a regular ring. The classification is in terms of the support of the pair (A,w) , denoted \supp(A,w) , where A=R[\chi_{1},\ldots,\chi_{n}] with w=f_{1}\chi_{1}+\cdots+f_{n}\chi_{n} and where each \chi_{i} has homological degree -2 ; see Definition 3.1 (as well as Proposition 3.2 ) for a precise definition of \supp(A,w) , and see Proposition 4.12 for a set theoretic description of it in this setting. Finally, let \V_{E}(M) denote the cohomological support of M over E as defined in [ 36 , 37 ] (inspired by the ones defined by Avramov and Buchweitz [ 1 , 3 ] over a complete intersection ring); cf. 4.6 . Theorem A . There is an inclusion preserving bijection \hbox to258.13pt{\vbox to28.85pt{\pgfpicture\makeatletter\hbox{\hskip 129.06467pt\lower-14.42696pt\hbox to0.0pt{\lxSVG@begingroup@{_scopebegin=1} \lxSVG@begingroup@{stroke=#000000} \lxSVG@begingroup@{fill=#000000} \lxSVG@setlinewidth{\the\pgflinewidth}\lxSVG@begingroup@{stroke-width=0.4pt} \lx@inpgf@ignorespaces\nullfont\lxSVG@begingroup@{_scopebegin=1} \lxSVG@closescope \hbox to0.0pt{\lxSVG@begingroup@{_scopebegin=1} {}{}{}{{}}\lx@inpgf@ignorespaces\hbox{\hbox{{\lxSVG@begingroup@{_scopebegin=1} {{}{}{{{}}{{}}}{{{\lx@inpgf@ignorespaces}}}{{}{}{{ {}{}}}{ {}{}} {{}{{\lx@inpgf@ignorespaces}}}{{}{\lx@inpgf@ignorespaces}}{}{{}{\lx@inpgf@ignorespaces}} {\lx@inpgf@ignorespaces }{{{{\lx@inpgf@ignorespaces}}\lxSVG@begingroup@{_scopebegin=1} \lxSVG@transformcm{1.0}{0.0}{0.0}{1.0}{-129.06467pt}{-2.5pt}\lxSVG@begingroup@{transform=matrix(1.0 0.0 0.0 1.0 -178.59 -3.46)} \pgfsys@hbox{62}\lxSVG@closescope }}}{{{\lx@inpgf@ignorespaces{}}}{{}}{{}}}} \lxSVG@closescope }}} {}{ {}{}{}}{}{ {}{}{}}{}{ {}{}{}} {{{{{}}{ {}{}}{}{}{{}{}}}}}{}{{{{{}}{ {}{}}{}{}{{}{}}}}}{{}}{}{}{}{}{}{{{}{}}}{}{}{{{}{}}}\lx@inpgf@ignorespaces\hbox{\hbox{\hbox{{\lxSVG@begingroup@{_scopebegin=1} {{}{{}}{}{}} \lxSVG@closescope }}}\hbox{{\lxSVG@begingroup@{_scopebegin=1} {{}{{}}{}{}} \lxSVG@closescope }}} {{}}{}{{}}{}{{{{}}}{{{}}}\lx@inpgf@ignorespaces}{{}}{{{\lx@inpgf@ignorespaces}}{{}}}{}{{\lx@inpgf@ignorespaces}}\lx@inpgf@ignorespaces\hbox{\hbox{{\lxSVG@begingroup@{_scopebegin=1} {{}{{}}{}{}} \lxSVG@closescope }}} {{}}{{{\lx@inpgf@ignorespaces}}{{}}}{}{{\lx@inpgf@ignorespaces}}\lx@inpgf@ignorespaces\hbox{\hbox{{\lxSVG@begingroup@{_scopebegin=1} {{}{{}}{}{}} \lxSVG@closescope }}} {{}}{}{{}}{}{{}} {}{}{}{}{{{}{}}}{}{{\lx@inpgf@ignorespaces}}{}{}{}{{{}{}}}\lxSVG@begingroup@{_scopebegin=1} \lxSVG@setlinewidth{\the\pgflinewidth}\lxSVG@begingroup@{stroke-width=0.39998pt} \lx@inpgf@ignorespaces{}{{ { \lxSVG@setdash{0.0pt}\lx@inpgf@ignorespaces\lxSVG@roundcap\lx@inpgf@ignorespaces\lxSVG@roundjoin\lx@inpgf@ignorespaces{} {\lx@inpgf@ignorespaces}{}{} {}{}{} \lxSVG@stroke}} }{}{}{{}}{}{}{{}}\lxSVG@stroke\lxSVG@drawpath@unclipped{M -21.88 4.77 L 10.96 4.77}{fill:none} {{}{{}}{}{}{{}}{{{\lx@inpgf@ignorespaces}}}}{{}{{}}{}{}{{}}{{{\lx@inpgf@ignorespaces}}{{{\lx@inpgf@ignorespaces}}{\lxSVG@begingroup@{_scopebegin=1} \lxSVG@transformcm{1.0}{0.0}{0.0}{1.0}{8.11777pt}{3.44443pt}\lxSVG@begingroup@{transform=matrix(1.0 0.0 0.0 1.0 11.23 4.77)} \lxSVG@begingroup@{_scopebegin=1} \lxSVG@begingroup@{stroke-dasharray=none,stroke-dashoffset=0.0pt} \lxSVG@begingroup@{stroke-linecap=round} \lxSVG@begingroup@{stroke-linejoin=round} \lxSVG@drawpath@unclipped{M -2.88 3.32 C -2.35 1.33 -1.18 0.39 0 0 C -1.18 -0.39 -2.35 -1.33 -2.88 -3.32}{fill:none} \lxSVG@closescope \lxSVG@closescope }}{{\lx@inpgf@ignorespaces}}}}\lx@inpgf@ignorespaces\hbox{\hbox{{\lxSVG@begingroup@{_scopebegin=1} {{}{}{{ {}{}}}{ {}{}} {{}{{\lx@inpgf@ignorespaces}}}{{}{\lx@inpgf@ignorespaces}}{}{{}{\lx@inpgf@ignorespaces}} {\lx@inpgf@ignorespaces }{{{{\lx@inpgf@ignorespaces}}\lxSVG@begingroup@{_scopebegin=1} \lxSVG@transformcm{1.0}{0.0}{0.0}{1.0}{-6.0658pt}{5.7972pt}\lxSVG@begingroup@{transform=matrix(1.0 0.0 0.0 1.0 -8.39 8.02)} \pgfsys@hbox{62}\lxSVG@closescope }}} \lxSVG@closescope }}} \lxSVG@closescope {}{ {}{}{}}{}{ {}{}{}}{}{ {}{}{}} {{{{{}}{ {}{}}{}{}{{}{}}}}}{}{{{{{}}{ {}{}}{}{}{{}{}}}}}{{}}{}{}{}{}{}{{{}{}}}{}{}{{{}{}}}\lx@inpgf@ignorespaces\hbox{\hbox{\hbox{{\lxSVG@begingroup@{_scopebegin=1} {{}{{}}{}{}} \lxSVG@closescope }}}\hbox{{\lxSVG@begingroup@{_scopebegin=1} {{}{{}}{}{}} \lxSVG@closescope }}} {{}}{}{{}}{}{{{{}}}{{{}}}\lx@inpgf@ignorespaces}{{}}{{{\lx@inpgf@ignorespaces}}{{}}}{}{{\lx@inpgf@ignorespaces}}\lx@inpgf@ignorespaces\hbox{\hbox{{\lxSVG@begingroup@{_scopebegin=1} {{}{{}}{}{}} \lxSVG@closescope }}} {{}}{{{\lx@inpgf@ignorespaces}}{{}}}{}{{\lx@inpgf@ignorespaces}}\lx@inpgf@ignorespaces\hbox{\hbox{{\lxSVG@begingroup@{_scopebegin=1} {{}{{}}{}{}} \lxSVG@closescope }}} {{}}{}{{}}{}{{}} {}{}{}{}{{{}{}}}{}{{\lx@inpgf@ignorespaces}}{}{}{}{{{}{}}}\lxSVG@begingroup@{_scopebegin=1} \lxSVG@setlinewidth{\the\pgflinewidth}\lxSVG@begingroup@{stroke-width=0.39998pt} \lx@inpgf@ignorespaces{}{}{}{}{{}}{}{}{{}}\lxSVG@stroke\lxSVG@drawpath@unclipped{M 11.51 -4.77 L -21.33 -4.77}{fill:none} {{}{{}}{}{}{{}}{{{\lx@inpgf@ignorespaces}}}}{{}{{}}{}{}{{}}{{{\lx@inpgf@ignorespaces}}{{{\lx@inpgf@ignorespaces}}{\lxSVG@begingroup@{_scopebegin=1} \lxSVG@transformcm{-1.0}{0.0}{0.0}{-1.0}{-15.6154pt}{-3.44443pt}\lxSVG@begingroup@{transform=matrix(-1.0 0.0 0.0 -1.0 -21.61 -4.77)} \lxSVG@begingroup@{_scopebegin=1} \lxSVG@begingroup@{stroke-dasharray=none,stroke-dashoffset=0.0pt} \lxSVG@begingroup@{stroke-linecap=round} \lxSVG@begingroup@{stroke-linejoin=round} \lxSVG@drawpath@unclipped{M -2.88 3.32 C -2.35 1.33 -1.18 0.39 0 0 C -1.18 -0.39 -2.35 -1.33 -2.88 -3.32}{fill:none} \lxSVG@closescope \lxSVG@closescope }}{{\lx@inpgf@ignorespaces}}}}\lx@inpgf@ignorespaces\hbox{\hbox{{\lxSVG@begingroup@{_scopebegin=1} {{}{}{{ {}{}}}{ {}{}} {{}{{\lx@inpgf@ignorespaces}}}{{}{\lx@inpgf@ignorespaces}}{}{{}{\lx@inpgf@ignorespaces}} {\lx@inpgf@ignorespaces }{{{{\lx@inpgf@ignorespaces}}\lxSVG@begingroup@{_scopebegin=1} \lxSVG@transformcm{1.0}{0.0}{0.0}{1.0}{-5.59274pt}{-8.81108pt}\lxSVG@begingroup@{transform=matrix(1.0 0.0 0.0 1.0 -7.74 -12.19)} \pgfsys@hbox{62}\lxSVG@closescope }}} \lxSVG@closescope }}} \lxSVG@closescope \lxSVG@closescope {{ {}{}{}}}{\lx@inpgf@ignorespaces}{\lx@inpgf@ignorespaces}\hss}\lxSVG@discardpath\lxSVG@closescope \hss}}\lxSVG@closescope\endpgfpicture}}\,, where the inverse bijections are given by \sigma(\mathsf{T})=\bigcup_{M\in\mathsf{T}}\V_{E}(M)\quad\text{and}\quad\tau(\mathcal{V})=\{M\in\Df(E)\mid\V_{E}(M)\subseteq\mathcal{V}\}\,. Specializing the theorem to when {\bm{f}} is an R -regular sequence recovers the classification of Stevenson in [ 40 ] for a complete intersection ring in Corollary 4.4 . In the case {\bm{f}}=0 , we have that E is an exterior algebra over a regular ring and Theorem A establishes a bijection between thick subcategories of \Df(E) and the specialization closed subsets of \spec(A); see Corollary 4.3 . There are a number of consequences of Theorem A on the homological algebra over E . For example, whenever \supp(A,w) is closed there is a classical generator for \Df(E) ; cf. Corollary 4.5 . This applies when {\bm{f}}=0 , or when {\bm{f}}\neq 0 and, for example, R is further assumed to be quasi-excellent and local. In a different direction, recall that E is a Gorenstein dg algebra in the sense of [ 24 ] , and hence, (-)^{\vee}=\RHom_{E}(-,E) is a self (quasi-)inverse auto-equivalence of \Df(E) . Corollary B . For each thick subcategory \mathsf{T} of \Df(E) , there is an equality \mathsf{T}=\mathsf{T}^{\vee} . For complete intersection rings this was established in [ 40 ] via Stevenson’s classification, and later also shown in [ 31 ] using a local-to-global principle from [ 13 ] . This also shows symmetric complexity holds over E ; recall the complexity of a pair M,N in \Df(E) measures the polynomial rate of growth of the minimal number of generators of \Ext_{E}^{n}(M,N) as n\to\infty . A consequence of Corollary B (given in Corollary 4.10 ) is that \cx_{E}(M,N)=\cx_{E}(N,M) , surprising (of course) because of the highly non-symmetric behavior of \Ext(-,-). The results above build on the philosophy from [ 2 , 25 , 37 , 44 ] that the nice homological properties of a complete intersection ring are merely inherited from the well-behaved homological properties of the Koszul complex over a regular ring. The proof of Theorem A , presented in Section 4 , depends on two major ingredients. The first is a curved version of the BGG correspondence from Martin’s thesis [ 32 ] . Namely, this establishes an equivalence of triangulated categories between \Df(E) and the category of perfect curved dg A

Read the complete text