The many faces of higher Hilbert spaces

This paper systematically unifies different notions of higher Hilbert spaces and their associated module categories by introducing $G$-dagger categories and $G$-Hermitian 2-vector spaces, where varying subgroups $G \leq O(2)$ recover distinct operator algebra structures like $\mathrm{C}^*$, $\mathrm{W}^*$, and $\mathrm{H}^*$-algebras, while also proposing criteria for positivity and an inductive framework for arbitrary dimensions.

The Gist

Imagine you are trying to build a house. In the world of standard mathematics, you have a very specific set of blueprints for a "Hilbert space" (a type of mathematical room used heavily in quantum physics). It's a room where you can measure distances and angles perfectly, and everything is "positive" (meaning distances are never negative).

Now, imagine you want to build a 2-story house (a "2-vector space"). You have the blueprints for the ground floor, but how do you build the second floor? The problem is that there isn't just one way to do it. Mathematicians have been arguing about the best way to build this second floor for a long time. Some say, "Let's just add a mirror!" (a dagger structure). Others say, "Let's add a special measuring tape!" (an inner product). Some say, "Let's do both!"

This paper, "The Many Faces of Higher Hilbert Spaces," is like a master architect stepping in to say: "Stop arguing. We can organize all these different blueprints into one single, unified system."

Here is how they do it, using some creative analogies:

1. The Compass and the Map (The O(2) Group)

The authors introduce a giant compass called O(2). Think of this compass as a set of rules for how you can rotate or flip your mathematical house.

• Flipping Bottom-to-Top ($Z^b_2$): Imagine flipping your house upside down. In math terms, this reverses the direction of the "rooms" (1-morphisms).
• Flipping Front-to-Back ($Z^t_2$): Imagine flipping the house so the front becomes the back. This reverses the direction of the "walls" or connections between rooms (2-morphisms).
• Rotating: You can also rotate the house.

The paper shows that every different way mathematicians have tried to define a "2-Hilbert space" corresponds to picking a specific subset of these compass directions.

• If you only allow Front-to-Back flips, you get what's called a $C^*$-category (a standard type of operator algebra).
• If you allow both flips, you get a $W^*$-category (a more complex type used in quantum field theory).
• If you allow everything (rotations and flips), you get a Baez 2-Hilbert space (the most "complete" version).

The paper draws a map (Diagram 1.1) showing how these different definitions are just different views of the same underlying structure, depending on which part of the compass you are looking at.

2. The "Positive" Test (Turning a Room into a Home)

Having a blueprint (a "Hermitian" structure) isn't enough. In the real world, you need a house to be "positive"—meaning it has a solid foundation and doesn't collapse. In math, this means your measurements must be positive numbers (you can't have a distance of -5 meters).

The authors propose a clever way to test if a 2-story house is "positive" without just guessing:

• The Elevator Test: Imagine sending a tiny elevator (a simple vector space) up into your 2-story house.
• The Reflection: You send the elevator up, bounce it off the ceiling (using the "dagger" or mirror), and bring it back down.
• The Result: If the elevator comes back as a "positive" object (a standard Hilbert space), then your whole 2-story house is a valid 2-Hilbert space.

This is the paper's "inductive" approach. Instead of defining the big house all at once, you check if the small parts inside it behave correctly. If every little piece you test turns out to be a "good" Hilbert space, then the whole structure is a "good" 2-Hilbert space.

3. The Translation to Algebra (The Language of Numbers)

The paper also translates these architectural ideas into the language of algebras (equations and numbers).

• They show that a "2-Hilbert space" is mathematically the same thing as a specific type of algebra called an $H^*$-algebra.
• They demonstrate that famous formulas used by physicists (like the "Connes fusion" formula) aren't magic tricks; they are just the natural result of following the rules of these compass flips and reflections.

The Big Picture

Think of the paper as a Rosetta Stone for higher mathematics.

• Before this paper, a mathematician might say, "I'm building a $C^*$-2-vector space," and another might say, "No, I'm building a Baez 2-Hilbert space," and they would think they were talking about two different things.
• This paper says, "You are both right. You are just using different settings on the same universal compass."

By organizing these definitions under the umbrella of G-dagger categories (categories with specific mirror/flip rules), the authors provide a systematic way to understand how these different mathematical structures relate to one another. They also suggest a recipe for building even taller "3-story" or "4-story" houses (higher Hilbert spaces) by using the same "elevator test" logic, ensuring that every level of the building is built on a solid, positive foundation.

In short: The paper takes a confusing mess of different definitions for "quantum rooms" and organizes them into a single, logical family tree based on how you can flip and rotate them, providing a clear recipe for building these structures in any dimension.

Technical Summary

Technical Summary: The Many Faces of Higher Hilbert Spaces

Problem Statement
The categorification of the notion of a Hilbert space into higher categories (specifically 2-vector spaces) has yielded multiple, seemingly distinct definitions in the literature. These include $C^*$-2-vector spaces, $W^*$-2-vector spaces, Baez 2-Hilbert spaces, and categories equipped with Calabi-Yau structures or non-degenerate pairings. While these structures are known to be related, there has been a lack of a systematic framework to organize, compare, and distinguish these notions, particularly regarding how "positivity" (the defining feature of Hilbert spaces) generalizes to higher dimensions. The paper addresses the need to unify these definitions and understand the precise relationship between different types of unitarity in finite-dimensional higher linear algebra.

Methodology
The authors employ the framework of $G$-dagger categories (where $G \le O(2)$ is a subgroup), building on recent developments in higher category theory [FHJF+24, M¨ul25, CL25]. The core methodology involves:

1. $O(2)$-volution and Fixed Points: The paper defines a twisted action of $O(2)$ on the 2-category of 2-vector spaces ($2\text{Vect}$), termed an $O(2)$-volution. This action combines the standard cobordism hypothesis action (involving duals and adjoints) with complex conjugation.
2. $G$-Hermitian Structures: By restricting this action to subgroups $G \le O(2)$, the authors define $G$-Hermitian 2-vector spaces as fixed points of the induced action on the core of $2\text{Vect}$.
   • $Z_2^b$ (bottom reflection) corresponds to reversing 1-morphisms (duals).
   • $Z_2^t$ (top reflection) corresponds to reversing 2-morphisms (opposite categories).
   • $SO(2)$ corresponds to pivotal structures (traces).
3. Selection of Dagger Objects: The authors propose that specific notions of "unitarity" or "Hilbert-ness" arise not just from the fixed-point structure itself, but from selecting specific classes of $G$-dagger objects. This selection process involves choosing "positive" fixed points, analogous to selecting positive-definite inner products in standard linear algebra.
4. Inductive Criteria: For subgroups containing $Z_2^b \times Z_2^t$ and $O(2)$, the paper proposes an inductive criterion for positivity. A $G$-2-vector space is defined as a $G$-Hilbert space if, for certain morphisms from the unit, the induced endomorphism (via the dagger structure) yields a "positive" object in the lower categorical dimension.

Key Contributions and Results

• Systematic Classification via Subgroups: The paper establishes a correspondence between subgroups of $O(2)$ and specific 2-categorical structures (Diagram 1.1 and Table 1):

  • $Z_2^b$: Corresponds to 2-vector spaces with a non-degenerate $\text{Vect}$-valued inner product ($2\text{Vect}_{ip}$).
  • $Z_2^t$: Corresponds to anti-involutive categories; restricting to "positive" fixed points yields $C^*$-2-vector spaces ($C^*2\text{Vect}$).
  • $Z_2^b \times Z_2^t$: Corresponds to $W^*$-2-vector spaces ($W^*2\text{Vect}$), combining inner products with $C^*$-structures.
  • $SO(2)$: Corresponds to Calabi-Yau categories ($CY2\text{Vect}$).
  • $O(2)$: Corresponds to Baez 2-Hilbert spaces ($2\text{Hilb}$), combining anti-involutive structures with positive definite traces.

• Unification of Operator Algebra Concepts: By translating these categorical definitions into the language of finite-dimensional algebras (via the equivalence between $2\text{Vect}$ and the Morita 2-category of algebras), the paper recovers fundamental constructions in operator algebra theory purely from categorical principles:

  • The formula for the operator-valued inner product on relative tensor products of Hilbert $C^*$-correspondences is derived from $Z_2^t$-fixed point data.
  • The Haagerup standard form of a von Neumann algebra emerges naturally as the canonical choice of $Z_2^b \times Z_2^t$-dagger object.
  • Connes fusion for bimodules is recovered as the composition of fixed-point data.

• Inductive Definition of Higher Hilbert Spaces: The paper outlines a conceptual, inductive procedure to define $n$-Hilbert spaces for arbitrary $n$. A $G$-dagger object in $n\text{Vect}$ is defined by requiring that the "inner product" of morphisms (viewed as objects in $(n-1)\text{Vect}$) lifts to an $(n-1)$-Hilbert space. This suggests a recursive structure where unitarity at level $n$ is determined by unitarity at level $n-1$.

Significance and Claims
The paper claims to provide a systematic framework that organizes the diverse landscape of 2-categorical unitarity. Its primary significance lies in demonstrating that the distinctions between $C^*$, $W^*$, and 2-Hilbert spaces are not arbitrary but correspond to specific choices of subgroups $G \le O(2)$ and specific selections of "positive" fixed points within those subgroups.

The authors modestly note that while they successfully define $G$-dagger objects for subgroups containing $Z_2^t$ (allowing for the definition of positivity via 2-morphisms), a fully satisfactory definition for subgroups not containing $Z_2^t$ remains an open problem, as the 2-morphisms (which form the vector space over $\mathbb{C}$) are necessary to define positivity.

Furthermore, the paper suggests that these finite-dimensional results serve as a blueprint for infinite-dimensional settings (e.g., infinite-dimensional $C^*$ and $W^*$-algebras) and higher dimensions ($n$-Hilbert spaces), though explicit generalizations to infinite dimensions require replacing duals with weaker notions, a direction the authors point to but do not fully resolve in this work. The paper concludes by conjecturing that this inductive approach correctly characterizes 3-Hilbert spaces as defined in recent literature.