Table of Contents
Fetching ...

Higher presentable categories and limits

Ko Aoki

TL;DR

This work develops a universe-free framework for presentable $n$-categories by introducing $\kappa$-compactly generated structures, defining $0\mathsf{Pr}^{\kappa}=\mathsf{Ani}$ and $(n+1)\mathsf{Pr}^{\kappa}=\mathrm{Mod}_{n\mathsf{Pr}^{\kappa}}\mathsf{Pr}^{\kappa}$, and forming $n\mathsf{Pr}=\varinjlim_{\kappa}n\mathsf{Pr}^{\kappa}$. It proves a positive adjoint-functor theorem for presentable $(\infty,2)$-categories: the underlying functor of any morphism has a right adjoint, hence these categories possess limits, while showing that the unit presentable $(\infty,3)$-category of presentable $(\infty,2)$-categories does not admit limits, answering Stefanich’s conjecture in the negative. The paper also develops enriched Ind-categories to model limits and conducts a detailed analysis of dominant morphisms, their complexity, and global limits, culminating in explicit nonpresentable pullbacks and mapping categories at higher levels. The results illuminate the asymmetry of adjoints and limits in higher category theory, provide a robust, enlargement-free construction of presentable $n$-categories, and reveal intrinsic higher-dimensional pathologies absent at the $(\infty,2)$-level.

Abstract

Stefanich generalized the notion of (locally) presentable $(\infty, 1)$-category to the notion of presentable $(\infty, n)$-category. We give a new description based on the new notion of $κ$-compactly generated $(\infty, n)$-category, which avoids universe enlargement. Using the new definition, we prove the underlying functor of a morphism between presentable $(\infty, 2)$-categories has a right adjoint. In particular, any presentable $(\infty, 2)$-category has limits. We also prove that this fails drastically when we go higher: The unit presentable $(\infty, 3)$-category, i.e., the category of presentable $(\infty, 2)$-categories, does not have limits. This settles Stefanich's conjecture in the negative.

Higher presentable categories and limits

TL;DR

This work develops a universe-free framework for presentable -categories by introducing -compactly generated structures, defining and , and forming . It proves a positive adjoint-functor theorem for presentable -categories: the underlying functor of any morphism has a right adjoint, hence these categories possess limits, while showing that the unit presentable -category of presentable -categories does not admit limits, answering Stefanich’s conjecture in the negative. The paper also develops enriched Ind-categories to model limits and conducts a detailed analysis of dominant morphisms, their complexity, and global limits, culminating in explicit nonpresentable pullbacks and mapping categories at higher levels. The results illuminate the asymmetry of adjoints and limits in higher category theory, provide a robust, enlargement-free construction of presentable -categories, and reveal intrinsic higher-dimensional pathologies absent at the -level.

Abstract

Stefanich generalized the notion of (locally) presentable -category to the notion of presentable -category. We give a new description based on the new notion of -compactly generated -category, which avoids universe enlargement. Using the new definition, we prove the underlying functor of a morphism between presentable -categories has a right adjoint. In particular, any presentable -category has limits. We also prove that this fails drastically when we go higher: The unit presentable -category, i.e., the category of presentable -categories, does not have limits. This settles Stefanich's conjecture in the negative.
Paper Structure (16 sections, 37 theorems, 39 equations)

This paper contains 16 sections, 37 theorems, 39 equations.

Key Result

Theorem A

With any enlargement of the universe, x2r4as coincides with Stefanich’s definition. In particular, his definition does not depend on the enlargement.

Theorems & Definitions (101)

  • Example 1.1
  • Definition 1.2
  • Theorem A
  • Conjecture 1.3: Stefanich
  • Remark 1.4
  • Remark 1.5
  • Example 1.6
  • Example 1.7
  • Theorem B
  • Remark 1.8
  • ...and 91 more