A Gauge-Invariant Clustering Coefficient for Complex-Weighted Bipartite Networks

summary

The gist

The gist: A gauge-invariant clustering coefficient for complex-weighted bipartite networks reduces to classical measures when phases vanish, becomes negative under destructive interference, and

In short

This work defines a new clustering coefficient for bipartite networks with complex edge weights, which are common in quantum and photonic systems where phases matter. The new coefficient reduces to the classical measure when phases are zero, becomes negative when alternative routes interfere destructively, and separates into topological and phase components. This shows that phase disorder increases the coherent distance while leaving hop-count distances unchanged.

Key concepts

Complex Edge Weights: In quantum or photonic systems, connections between nodes can have complex weights instead of just real numbers. These complex weights have both a magnitude (strength) and a phase. The phase determines how signals or paths interfere with each other when traveling through the network.
Bipartite Network Structure: A bipartite network is one where the nodes are divided into two distinct sets, say U and V, and connections only exist between nodes in different sets. In this structure, closed loops of three nodes (triangles) do not exist; instead, the smallest cycle is a four-node square.
Clustering Coefficient Factorization: The complex clustering coefficient ($C_u$) can be broken down into two distinct parts: a topological factor ($C^{(0)}_u$) and a phase factor ($Q_u$). The topological part depends only on the network's structure (like how many connections exist) and the phase part depends only on the fluxes (the accumulated phases around squares). This separation is key to analyzing how structure and phase disorder affect clustering.

Terminology used across episodes

Reinforcement Learning (RL) — explained in 25 episodes
Federated Learning — explained in 16 episodes
Retrieval-Augmented Generation (RAG) — explained in 16 episodes
Interpretability — explained in 15 episodes
Sample Complexity — explained in 15 episodes
Generalization — explained in 13 episodes

This episode discusses

A Gauge-Invariant Clustering Coefficient for Complex-Weighted Bipartite Networks · Paper Radio

The paper

A Gauge-Invariant Clustering Coefficient for Complex-Weighted Bipartite Networks · Read on arXiv

Anjali Waghmare, Yu Tian, Renaud Lambiotte

Mathematical Institute, University of Oxford · Center for Systems Biology Dresden

Structural measures such as the clustering coefficient and the average shortest-path length characterise how a network is organised. These measures are typically formulated for real, non-negative edge weights. A class of quantum and photonic architectures has complex edge weights instead, whose phases determine whether alternative routes interfere constructively or destructively. These architectures are also bipartite, so triangles are absent and the smallest closed cycle is a four-node square. Existing measures address complex weights and bipartite structure separately: bipartite clustering coefficients quantify clustering through four-node squares but do not contain phase information, while interferometric coefficients retain phase but are defined on triangles. In this work, we define a clustering coefficient for complex-weighted bipartite networks which reduces to the classical bipartite coefficient when the phases vanish, becomes negative when alternative routes cancel, and can be obtained for all nodes from a single sparse matrix product. We show that the phase accumulated around a square is the smallest gauge-invariant carrier of structural phase information in a bipartite network. The clustering coefficient factorises into a topological contribution and a phase contribution, making a bipartite Watts--Strogatz ensemble analytically tractable. For phases uniformly distributed on [-Δ,Δ], the mean phase contribution is (Δ/Δ) 4, independent of node, degree, and topology. We also derive closed-form expressions for the clustering coefficient, open-path visibility, and phase variance. We numerically verify these predictions and further show that phase disorder increases coherent distance.

Transcript

Introduction to the show: ident: Quantum Radio. Generated commentary on the latest quantum physics and condensed matter papers.

Kai: I'm Kai, and with me are Mira and Lev, guest researcher.

Mira: Today's paper: "A Gauge-Invariant Clustering Coefficient for Complex-Weighted Bipartite Networks".

Kai: The gist: A gauge-invariant clustering coefficient for complex-weighted bipartite networks reduces to classical measures when phases vanish, becomes negative under destructive interference, and factorizes into topological and phase contributions.

Mira: First, who's behind it and why it matters.

Title and authors: Mira: We just touched on the title and authors of this paper, "A Gauge-Invariant Clustering Coefficient for Complex-Weighted Bipartite Networks." The core idea is that they are providing a structure measure that works for networks where the connections aren't just simple real numbers but have phases attached to them.

Kai: What that means in plain language is they’re creating a way to measure how connected things are, but this measurement remains consistent even when those phases shift around, as long as you keep the right mathematical rules.

Lev: From an error correction standpoint, that gauge invariance is crucial because if we can't define a structure measure that ignores the arbitrary phase choices you make when setting up your system, then any measurement of connectivity becomes meaningless.

Kai: Right, so it’s about making sure the structural information you get from measuring the network doesn't just depend on an arbitrary choice of how you defined those complex weights.

The paper's summary: Mira: The summary explains that this new clustering coefficient actually reduces to the classical one we already know, which is what happens when all those phases just vanish or become zero.

Kai: It’s like if you have a quantum circuit where everything is perfectly aligned and the phases are zero everywhere, this new complex measure just spits out the same classical result you get from looking only at the link strengths.

Lev: But what they’re emphasizing is that it also gets negative when there's destructive interference between alternative routes, which is a key physical effect we see in quantum transport.

Mira: And they show this coefficient factorizes into two parts: a topological part that only cares about the shapes and sizes of the connections, and a phase part that only depends on the fluxes around those squares.

The paper's improvements: Kai: The paper suggests a big improvement by proving this factorization, meaning you can separate what’s purely about the geometry from what’s about the interference effects.

Mira: That allows us to study the structural organization—the topology—completely independently of how messy those phases are, and then analyze the phase part separately based on just the fluxes.

Lev: For experimentalists, that factorization is pretty helpful because you can measure one part of it to check if your physical setup is behaving correctly structurally, and another part to see how much interference you're actually getting.

Kai: So they're moving from just measuring a single number for the whole network structure to having two distinct pieces that tell us different things about the network organization.

Conclusion: Mira: To wrap up, the paper shows that this gauge-invariant clustering coefficient is robust because it keeps the topological factor tied only to connectivity, while the phase factor handles the interference effects based on fluxes.

Kai: It’s interesting how they found that even when you have complete phase disorder—meaning every link has a random phase between negative and positive values—the mean phase contribution is still predictable, specifically (sin ∆/∆)four <ref:2610.11929#pg2>.

Lev: What this means for us running experiments is that while the hop-count distance might not change much with random phases, the coherent distance does increase by sixteen percent on networks with the same connectivity.

Mira: And they also point out a limitation: this specific treatment assumes phases are drawn independently for each link and that they are static, meaning it’s a steady-state analysis rather than looking at dynamic changes.

Kai: So basically, the paper gives us a tool to see structure through complex weights without getting lost in every single interference detail, and we have to remember that their results hold under those specific assumptions about phase independence.

More episodes

← Home