Derived projective covers and Koszul duality of simple-minded and silting collections

This paper introduces derived projective covers to establish an if-and-only-if criterion for them forming a silting collection and proves a Koszul duality result linking silting and simple-minded collections.

The Gist

Imagine you are trying to understand a massive, chaotic library of mathematical objects. This library is called a Triangulated Category. It's full of complex shapes, structures, and relationships that are hard to see all at once.

To make sense of this library, mathematicians use two main "organizing systems" (or lenses):

1. The "Time" Lens ($t$-structures): This organizes objects based on their "history" and "future." It helps us find the "simplest" building blocks, which the author calls Simple-Minded Collections. Think of these as the pure, indivisible atoms of the library.
2. The "Weight" Lens (Weight structures): This organizes objects based on their "heaviness" or complexity. It helps us find the "strongest" building blocks, called Silting Collections. Think of these as the heavy-duty scaffolding or the indestructible pillars that hold everything up.

For a long time, mathematicians knew these two lenses were related, but the connection was a bit fuzzy. This paper by Lukas Bonfert acts like a master key that unlocks the precise relationship between them.

Here is the breakdown of the paper's main ideas using everyday analogies:

1. The "Derived Projective Cover": The Perfect Blueprint

Imagine you have a fragile, simple object (a "Simple-Minded" atom). You want to build a sturdy, complex structure (a "Silting" pillar) that perfectly supports it.

• The Problem: Sometimes, you can't just build a pillar directly. You need a "cover" or a blueprint that wraps around the simple object to make it strong enough to be a pillar.
• The Solution: Bonfert introduces the concept of a Derived Projective Cover.
  • Think of a Simple Object as a raw diamond.
  • Think of a Silting Object as a diamond set in a heavy, indestructible gold frame.
  • The Derived Projective Cover is the process of taking that raw diamond and finding the exact gold frame that fits it perfectly.
• The Big Discovery: The paper proves a "Golden Rule": You can perfectly match every simple atom to a heavy pillar if and only if every simple atom has a perfect "cover" (a gold frame) available. If you can find these covers, the two organizing systems (Time and Weight) are perfectly aligned.

2. The "Koszul Duality": The Mirror World

This is the most magical part of the paper. In mathematics, there is a concept called Duality, which is like looking in a mirror. What is small on one side looks big on the other; what is simple on one side looks complex on the other.

• The Analogy: Imagine two different languages.
  • Language A speaks in "Simple Atoms" (Simple-Minded Collections).
  • Language B speaks in "Heavy Pillars" (Silting Collections).
• The Discovery: Bonfert shows that these two languages are actually Koszul Duals of each other. This means:
  • If you take the "dictionary" (the algebraic structure) of the Simple Atoms and translate it through a special mirror (Koszul Duality), you get the dictionary of the Heavy Pillars.
  • Conversely, if you start with the Heavy Pillars and look in the mirror, you get the Simple Atoms.
• Why it matters: It means you don't have to study both systems separately. If you understand one, the mirror tells you everything about the other. It's like realizing that the blueprint for a house and the finished house are actually two sides of the same coin.

3. The "Natural" Connection

Finally, the paper shows that this relationship isn't just a lucky accident for one specific library. It is natural.

• The Analogy: Imagine you have a map of a city (the library). If you zoom in or zoom out (change the category), or if you translate the map from English to Spanish (apply a functor), the relationship between the "Simple Atoms" and the "Heavy Pillars" stays consistent. The "mirror" doesn't crack; it just scales up or down. This proves the theory is robust and works everywhere, not just in isolated cases.

Summary: What did we learn?

1. The Bridge: We found a precise way to connect "Simple" things (atoms) with "Strong" things (pillars) using a concept called a Derived Projective Cover.
2. The Condition: This connection works perfectly only when every simple thing has a matching strong cover.
3. The Mirror: These two worlds are Koszul Duals. They are mirror images of each other. Understanding one automatically gives you the code to understand the other.
4. The Result: This unifies several different areas of mathematics (like algebra and geometry) under one single, elegant framework.

In a nutshell: The paper tells us that the "simplest" parts of a mathematical universe and the "strongest" parts are actually two sides of the same coin, connected by a perfect matching system and reflected in a magical mirror.

Technical Summary

1. Problem Statement

The paper addresses the structural relationships within triangulated categories, specifically focusing on the interplay between $t$-structures (associated with simple-minded collections) and weight structures (associated with silting collections).

Key problems tackled include:

• Characterizing Orthogonality: While bijections between bounded $t$-structures with finite-length hearts and bounded weight structures with Krull–Schmidt cohearts are known (e.g., via Koenig-Yang), the precise conditions under which these structures are orthogonal (specifically $w$-$t$-strict orthogonality) require deeper characterization using derived projective objects.
• Existence of Derived Projective Covers: Determining when a triangulated category admits "enough derived projectives" relative to a $t$-structure, and establishing a criterion for when derived projective covers of simple objects form a silting collection.
• Koszul Duality: Formalizing a Koszul duality between simple-minded collections and silting collections. While classical Koszul duality swaps simples and indecomposable projectives in abelian categories, this paper seeks to extend this duality to the derived setting, swapping simple-minded collections and silting collections via dg-algebra Koszul duality.
• Naturality: Investigating whether the bijections between these structures are natural with respect to adjoint functors.

2. Methodology

The author employs a combination of homological algebra, triangulated category theory, and dg-category techniques:

• Derived Projective Objects: The paper introduces and utilizes the concept of derived projective objects (objects $P$ in a $t$-structure heart such that $\text{Hom}(P, X[1])=0$ for $X$ in the heart). This generalizes the notion of projective objects in abelian categories to the triangulated setting.
• Derived Projective Covers: A new concept is defined where a morphism $\pi: P \to X$ is a derived projective cover if $P$ is derived projective and its truncation $H^0_t(\pi)$ is a projective cover in the heart.
• Orthogonality Levels: The paper refines the definitions of orthogonality between weight structures ($w$) and $t$-structures ($t$), distinguishing between $w$-strict, $t$-strict, and $w$-$t$-strict orthogonality.
• DG-Enhancements and Koszul Duality: The author uses dg-enhanced triangulated categories and the dg-Koszul duality framework (following Keller) to relate the endomorphism dg-algebras of silting collections and simple-minded collections.
• ST/WT Pairs: The paper revisits "ST pairs" (Silting-Simple-Minded) from previous literature, renaming them WT pairs (Weight-T-structure) to emphasize that the correspondence can be defined purely via weight and $t$-structures without explicit reference to silting collections.

3. Key Contributions and Results

A. Derived Projective Covers and Silting Collections (Section 3)

The paper establishes a fundamental equivalence (Theorem A / Theorem 3.16) linking four concepts for a non-degenerate $t$-structure $t$ with a finite-length heart:

1. $t$ is a silting $t$-structure.
2. There exists a bijection $\phi$ between the set of indecomposable derived projectives $\mathcal{P}$ and simple objects $\mathcal{L}$ satisfying a specific orthogonality condition: $\text{Hom}(P, L[m]) \cong \text{End}(L)$ if $L=\phi(P)$ and $m=0$, and $0$ otherwise.
3. Every simple object in the heart admits a derived projective cover, and $\mathcal{P}$ is exactly the set of these covers.
4. The category has enough derived projectives with respect to $t$.

Significance: This result provides a necessary and sufficient criterion for a $t$-structure to be silting, framed entirely in terms of the existence of derived projective covers for simple objects. It generalizes results from [CSPP22] and [KY14].

B. The WT Correspondence (Section 3.4)

The paper formalizes the correspondence between weight structures and $t$-structures via WT pairs.

• Theorem 3.22: It proves that for a WT pair, the bijection between bounded weight structures on a subcategory $\mathcal{C}$ and bounded $t$-structures on $\mathcal{D}$ is given precisely by $w$-$t$-strict orthogonality.
• This unifies previous results (e.g., Koenig-Yang for finite-dimensional algebras, results for non-positive dg algebras) under a single framework where the correspondence is mediated by derived projective covers and simple tops.

C. Koszul Duality (Section 4)

The paper proves a Koszul duality theorem (Theorem 4.6) for compact silting collections and simple-minded collections in dg-enhanced triangulated categories:

• Let $\mathcal{P}$ be a compact silting collection and $\mathcal{L}$ the set of simple objects in the associated silting heart.
• The dg-algebra $\text{End}(\bigoplus_{L \in \mathcal{L}} L)$ is the dg-Koszul dual of $\text{End}(\bigoplus_{P \in \mathcal{P}} P)$.
• Conversely, under finite-dimensionality assumptions on cohomology, the endomorphism algebra of the silting collection is the Koszul dual of the simple-minded collection.
• Examples: The author provides explicit computations for the $A_2$ quiver, demonstrating how standard and non-standard mutations of collections affect the degrees and differentials in the resulting dg-algebras, confirming the duality.

D. Naturality of Orthogonality (Section 5)

The paper proves that $w$-$t$-strict orthogonality is natural with respect to adjoint functors (Theorem 5.1).

• If $F: \mathcal{C} \to \mathcal{C}'$ and $G: \mathcal{D}' \to \mathcal{D}$ are adjoint functors (in a dual sense), then $F$ is weight-exact if and only if $G$ is $t$-exact.
• This implies that the bijection between weight structures and $t$-structures is preserved under these functors, reinforcing the robustness of the correspondence.

4. Significance and Impact

• Unification: The paper unifies various scattered results regarding the correspondence between $t$-structures and weight structures (from finite-dimensional algebras to dg-algebras) under the umbrella of "WT pairs" and derived projective covers.
• Conceptual Clarity: By introducing "derived projective covers," the author provides a concrete mechanism to understand the link between silting objects and simple objects, analogous to the classical link between projective covers and simple modules in abelian categories.
• Koszul Duality Extension: It successfully extends the classical Koszul duality (which swaps simples and projectives) to the derived setting, swapping simple-minded collections and silting collections. This offers a powerful tool for studying the derived categories of dg-algebras and their mutations.
• Technical Rigor: The paper resolves ambiguities regarding the strictness of orthogonality conditions, showing that in many relevant cases (adjacent structures), the various levels of strictness coincide, simplifying the verification of these correspondences.

In summary, Bonfert's work provides a comprehensive theoretical framework connecting silting theory, $t$-structures, and Koszul duality, utilizing derived projective covers as the central bridge between these concepts.