Imagine you have a group of friends, and you want to know if they can all agree on a single "secret language" (a common basis) to communicate without any confusion. In the quantum world, this "secret language" is a specific way of looking at a system where everything is clear and diagonal (no hidden overlaps).

If your group of friends (quantum states) can all speak this same secret language, they are "set incoherent" (they get along perfectly). If they cannot agree on one language and are constantly talking past each other, they are "set coherent" (they have a relational quantum resource).

The problem is: You aren't allowed to see their actual faces or hear their voices directly. You can only ask them to perform specific, tricky math tricks involving their own reflections and overlaps. These math tricks are called Bargmann invariants.

This paper asks a simple question: How many of these math tricks do we need to perform to know for sure if the group can agree on a secret language?

Here is the hierarchy the authors discovered, explained with everyday analogies:

1. The "Two-Headed" Test (Qubits / 2 Dimensions)

Imagine you have two people. To see if they can agree on a language, you check two things:

How "pure" or distinct each person is individually.
How much they overlap when they stand next to each other.

The Result: For two people in a simple 2-dimensional world (like a coin flip, heads or tails), checking these two things is enough. If the math works out a certain way, you know they can agree on a language. If not, they can't. It's like checking if two arrows are pointing in the exact same line; if they are, they are compatible.

2. The "Three-Headed" Test (Qutrits / 3 Dimensions)

Now, imagine the world gets slightly more complex (3 dimensions). You still have two people, but they have more ways to move.

The 2nd Test Fails: Checking just their individual purity and overlap isn't enough anymore. They might look compatible on the surface but have hidden disagreements in their third dimension.
The 3rd Test Works: If you add a third layer of math (looking at how they interact in a specific 3-step sequence), you can finally tell if they agree. In this 3D world, knowing their "shape" (spectrum) and how they twist around each other is enough to solve the puzzle.

3. The "Four-Headed" Trap (4 Dimensions and up)

The world gets even bigger (4 dimensions).

The 3rd Test Fails Again: Even if you check all the 3-step interactions, you can still be fooled! The authors found a clever example where two groups of states look identical in every 3-step test, but one group actually agrees on a language while the other is secretly fighting.
The Lesson: In higher dimensions, looking at "how much they overlap" and "how they twist in 3 steps" is not enough to catch the disagreement.

4. The Universal "Order-Sensitive" Test (The 4th Order Solution)

The authors found the ultimate solution that works for any size group, no matter how complex the dimension is.

They realized that to catch the disagreement, you need to check the order in which things happen.

Imagine two people, Alice and Bob.
Test A: Alice speaks, then Bob speaks, then Alice speaks, then Bob speaks ( $A \to B \to A \to B$ ).
Test B: Alice speaks, then Alice speaks, then Bob speaks, then Bob speaks ( $A \to A \to B \to B$ ).

In a world where everyone agrees on a language, the order doesn't matter; the result is the same. But if they are fighting (non-commuting), the order does matter.

The Breakthrough: The paper proves that the difference between these two specific 4-step sequences is a perfect, universal detector.

If the difference is zero, they can agree on a secret language.
If the difference is anything else, they cannot.

Summary of the Hierarchy

The paper builds a ladder of complexity to solve this puzzle:

Level 2 (Simple): Works for 2D pairs. (Like checking if two arrows are parallel).
Level 3 (Medium): Works for 3D pairs. (Like checking the shape and twist of 3D objects).
Level 4 (Universal): Works for everything. It detects the "non-commutativity" (the fighting) by comparing the order of operations.

Why This Matters

The authors show that you don't need to know the full, complicated details of the quantum states to know if they are compatible. You just need to run these specific, low-level math "tricks" (Bargmann invariants).

For small groups (2D): A simple check is enough.
For medium groups (3D): You need a slightly deeper check.
For big groups (4D+): You must check the order of events (the 4th-order test) to be absolutely sure.

This provides a "low-order hierarchy," meaning we can stop looking for more complex data once we reach the 4th order. It's a complete, basis-independent rulebook for deciding if a family of quantum states can ever agree on a common language.

Technical Summary: A Low Order Bargmann Invariant Hierarchy for Set Coherence

Problem Statement

Quantum coherence is typically defined relative to a fixed reference basis. However, a single density operator possesses no intrinsic, basis-independent coherence as it can always be diagonalized in its own eigenbasis. A nontrivial, basis-independent notion arises when considering a finite family of quantum states $\vec{\rho} = (\rho_1, \dots, \rho_n)$ . This property, termed set coherence, exists if the states cannot be simultaneously diagonalized in a common basis. Conversely, the family is set incoherent if such a common basis exists, which is mathematically equivalent to the pairwise commutativity of all states in the family ( $[\rho_i, \rho_j] = 0$ ).

The central problem addressed in this work is determining the minimal amount of basis-independent data required to decide whether a family of states is set coherent. Specifically, the authors investigate how low the "order" of Bargmann invariants (cyclic traces of products of states) must be to fully characterize set coherence across different Hilbert space dimensions ( $d$ ). The goal is to identify the lowest-order data that can distinguish between commuting (set incoherent) and non-commuting (set coherent) families without reconstructing the density matrices or choosing a reference basis.

Methodology

The paper employs the framework of Bargmann scenarios, where a scenario $W$ is a finite set of words (sequences of state labels) defining a set of generalized Bargmann invariants:
$\Delta_w(\vec{\rho}) = \text{Tr}(\rho_{i_1}\rho_{i_2}\cdots\rho_{i_m})$
These invariants are invariant under simultaneous unitary conjugation. The authors analyze the completeness of specific scenarios $W$ in dimension $d$ . A scenario is complete if the set of invariants generated by set-incoherent families ( $C_d(W)$ ) is disjoint from the set generated by set-coherent families ( $I_d(W)$ ).

The analysis proceeds by constructing a hierarchy of scenarios based on the order of the words (the number of states in the trace product):

Second-order ( $W_2$ ): Includes purities ( $\text{Tr}(\rho^2)$ ) and overlaps ( $\text{Tr}(\rho\sigma)$ ).
Third-order ( $W_3$ ): Includes all cyclic words of length $\le 3$ (e.g., $\text{Tr}(\rho^3)$ , $\text{Tr}(\rho^2\sigma)$ ).
Fourth-order ( $W_4$ ): Includes ordering-sensitive words of length 4, specifically comparing $\text{Tr}(\rho^2\sigma^2)$ and $\text{Tr}(\rho\sigma\rho\sigma)$ .

The authors construct explicit counterexamples (pairs of states with identical lower-order invariants but different commutativity properties) to demonstrate incompleteness and provide algebraic proofs for completeness in specific dimensions.

Key Contributions and Results

1. Dimension-Dependent Hierarchy for Two States

The paper establishes a strict hierarchy regarding the sufficiency of low-order data for two states $(\rho, \sigma)$ :

Qubits ( $d=2$ ): Second-order data ( $W_2$ ) are complete. The Hilbert-Schmidt lengths and overlap are sufficient to determine commutativity because, in the Bloch representation, commutativity is equivalent to the collinearity of Bloch vectors, which is fully determined by their lengths and inner product.
Qutrits ( $d=3$ ): Second-order data are incomplete. While they determine the spectra and overlap, they fail to capture the relative eigenbasis structure necessary to decide commutativity. However, third-order data ( $W_3$ ) are complete for qutrits. In dimension 3, the first three moments determine the characteristic polynomial (and thus the spectrum) of a density operator. The mixed moments in $W_3$ constrain the relative arrangement of eigenspaces sufficiently to enforce commutativity if the invariants match a commuting pair.
Dimension Four and Above ( $d \ge 4$ ): Third-order data are incomplete. The authors construct a counterexample where a commuting pair and a non-commuting pair share identical $W_3$ invariants. The failure persists for all $d \ge 4$ via direct sum embeddings.

2. Universal Fourth-Order Criterion

The primary theoretical contribution is the identification of a universal pairwise criterion valid for any finite dimension $d$ .

The scenario $W_4 = \{1122, 1212\}$ , consisting of the words $\text{Tr}(\rho^2\sigma^2)$ and $\text{Tr}(\rho\sigma\rho\sigma)$ , is complete for all $d$ .
The authors prove the identity:
$\Gamma(\rho, \sigma) := \text{Tr}(\rho^2\sigma^2) - \text{Tr}(\rho\sigma\rho\sigma) = \frac{1}{2} \| [\rho, \sigma] \|_2^2$
where $\| \cdot \|_2$ is the Hilbert-Schmidt norm.
Consequently, $\Gamma(\rho, \sigma) = 0$ if and only if $[\rho, \sigma] = 0$ .
This result extends to arbitrary finite families: a family is set incoherent if and only if $\Gamma(\rho_i, \rho_j) = 0$ for all pairs $i < j$ .

3. The Bargmann Gap

The quantity $\Gamma(\rho, \sigma)$ is termed the Bargmann gap. It serves as a faithful, basis-independent indicator of noncommutativity.

Spectral Interpretation: In the eigenbasis of $\rho$ , the gap is given by $\sum_{i<j} (\lambda_i - \lambda_j)^2 |\sigma_{ij}|^2$ . It detects exactly the coherences of $\sigma$ between eigenspaces of $\rho$ with different eigenvalues.
Noise Robustness: For depolarized states, the gap scales by a known multiplicative factor $(1-p)^2(1-q)^2$ , making the signal suppression transparent.

4. Extension to Many-State Families

The Appendix demonstrates that the limitations of lower-order data persist for families with $n > 2$ states.

Pairwise second-order data are insufficient for qutrit families ( $n \ge 2$ ).
Pairwise third-order data (all words of length $\le 3$ ) are insufficient for four-dimensional families.
This confirms that the fourth-order requirement is not an artifact of the two-state setting but a fundamental property of the relational resource.

Significance

The paper provides a rigorous low-order hierarchy connecting cyclic trace invariants to the noncommutativity that prevents a common incoherent basis.

It clarifies that while low-order data (2nd and 3rd order) can be sufficient in low dimensions due to specific spectral constraints (e.g., the determination of the characteristic polynomial by the first three moments in $d=3$ ), these mechanisms fail in higher dimensions.
It identifies fourth-order, ordering-sensitive invariants as the first universal pairwise criterion for set coherence.
The work offers a compact, basis-independent decision rule for set coherence that does not require full state tomography or reference frame selection, relying instead on quantities that are operationally accessible via multi-copy or interferometric protocols.

The authors note that while a related article on commutativity involving a single order-4 equality was recently identified, this work establishes the specific hierarchy and the "Bargmann gap" as a quantitative measure of the relational resource.

A low order Bargmann invariant hierarchy for set coherence