⚛️ quantum physics

Kostant relation in filtered randomized benchmarking for passive bosonic devices

This paper introduces two efficient filter functions based on immanants and special unitary group characters to reduce the computational cost and variance of bosonic randomized benchmarking for passive devices, enabling simpler platform characterization without requiring Clebsch-Gordan coefficients.

Original authors: David Amaro-Alcalá

Published 2026-04-13

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: David Amaro-Alcalá

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you have a very delicate, high-tech machine made of light (photons) passing through a maze of mirrors and beam splitters. This is a passive bosonic device, and it's a key building block for the next generation of quantum computers.

But here's the problem: these machines are fragile. They get noisy. Light gets lost, or sometimes extra light sneaks in. To know if your machine is working well, you need to run a "health check" called Randomized Benchmarking (RB).

Think of RB like a stress test for a car engine. You run the engine through thousands of random, complex routes to see how much it degrades over time. If the engine holds up, it's good. If it sputters, you know there's a problem.

The Old Problem: The "Math Nightmare"

The previous method for testing these light-machines had two major headaches:

The "Permanent" Problem: To analyze the data, the old method required calculating something called a matrix permanent. In the world of math, calculating a permanent is like trying to solve a Sudoku puzzle that grows exponentially harder with every extra number you add. It's so hard that even supercomputers struggle with it. It's like trying to count every possible way to arrange a deck of cards, but the deck keeps getting bigger.
The "Fock State" Problem: The experiment required preparing very specific, rare states of light (called Fock states) and using incredibly expensive, complex detectors that can count exactly how many photons hit them. It's like trying to test a car engine by only using a specific brand of premium fuel and a $100,000 diagnostic tool that most garages don't have.

The New Solution: The "Kostant Shortcut"

This paper introduces a new, smarter way to do the health check. The author, David Amaro-Alcalá, uses a mathematical "shortcut" discovered by a mathematician named Bertram Kostant.

Here is the analogy:
Imagine you are trying to measure the "vibe" of a crowded party.

The Old Way: You try to count every single conversation, every handshake, and every laugh individually. You need a massive spreadsheet and a team of accountants (the CG coefficients and permanents) to make sense of it all.
The New Way (Kostant Relation): You realize that you don't need to count every conversation. Instead, you just need to listen to the overall hum of the room (the "character" of the group) or look at specific, simple patterns (the "immanants").

The paper proposes two new "filters" to simplify the math:

1. The Character Filter (The "Magic Constant")

This is the star of the show. Instead of doing the impossible math of counting every arrangement, this filter uses the characters of the special unitary group.

Analogy: Think of a choir. Instead of trying to write down the exact note every single singer is hitting to understand the song, you just listen to the harmony. The harmony (the character) tells you everything you need to know about the song's quality, and it's incredibly easy to calculate.
The Benefit: This filter has a constant, low variance. In plain English, the results are super stable and predictable. It's like using a ruler that never bends, no matter how many times you use it. It's fast, easy to compute, and works for any size of the machine.

2. The Immanant Filter (The "Simplified Pattern")

This is a middle-ground option. It uses "immanants," which are like a simplified version of the hard math.

Analogy: If the old math was trying to solve a 100-piece puzzle, this filter helps you solve a 10-piece puzzle that still gives you the right picture. It removes the need for the most confusing parts of the math (the Clebsch-Gordan coefficients).

Making it Practical: "Good Enough" Light

The paper also solves the experimental headache.

Old Requirement: You needed perfect, single photons (like a single, perfect raindrop) and a detector that could count them one by one.
New Reality: The authors show you can use weak coherent states (basically, a very dim laser pointer) and simple intensity detectors (which just measure "how bright" the light is, not how many drops are in it).
Analogy: Instead of needing a high-speed camera to count individual raindrops to test a roof, you can just use a bucket to see how much water leaks through. It's "good enough" and much cheaper. Their computer simulations showed that even with this simpler equipment, the results were almost identical to the expensive, perfect setup.

Why This Matters

This paper is a game-changer because it turns a theoretical, super-hard problem into something a regular physics lab can actually do.

Cheaper: You don't need the most expensive detectors.
Faster: You don't need a supercomputer to crunch the numbers; a standard laptop can handle the "character filter."
Robust: It works even if the machine loses or gains a few photons (which happens all the time in the real world).

In summary: The author took a complex, expensive, and mathematically terrifying process for testing quantum light-machines and replaced it with a simple, stable, and cheap method using a clever mathematical shortcut. It's like swapping a manual transmission that requires a PhD to operate for an automatic transmission that anyone can drive.

1. Problem Statement

The paper addresses the computational and experimental limitations of the existing Bosonic Randomized Benchmarking (RB) protocol for characterizing passive bosonic devices (e.g., linear optical interferometers). The original proposal (Arienzo et al., 2025) faces two critical bottlenecks:

Computational Hardness: The original filter requires calculating matrix permanents, which are #P-hard to compute. Furthermore, these permanents depend on complex decompositions involving Clebsch–Gordan (CG) coefficients, making the data analysis analytically opaque and computationally expensive.
Experimental Inaccessibility: The protocol theoretically requires the preparation of specific Fock states (photon number states) and the use of photon-number-resolving (PNR) detectors, which are technologically challenging for many laboratories.

The goal is to develop a simplified filtering scheme that retains the robustness of RB (insensitivity to State Preparation and Measurement errors) while reducing computational overhead and relaxing experimental requirements.

2. Methodology

The author proposes a new filtering framework based on Representation Theory of the special unitary group $SU(m)$, leveraging two main mathematical tools: Immanants and Characters.

A. Theoretical Framework

Representation Decomposition: The Hilbert space of $n$ indistinguishable photons in $m$ modes corresponds to the symmetric irreducible representation (irrep) $\lambda$ of $SU(m)$. The noise channel acts on the tensor product $\lambda \otimes \lambda^*$ , which decomposes into a direct sum of irreps $\mu$ .
Kostant Relation: The core theoretical innovation utilizes a relation by Bertram Kostant. This relation states that the immanant of a unitary matrix $U$ (associated with a partition $\kappa$ ) can be computed as the sum of diagonal matrix elements (D-functions) over the zero-weight states of that irrep:
$\sum_{|\zeta_\kappa\rangle \in Z_\kappa} \langle \zeta_\kappa | \Gamma(U) | \zeta_\kappa \rangle = \text{Imm}_\kappa(U)$
This allows the filter to bypass the explicit calculation of CG coefficients and projectors.

B. Proposed Filters

The paper introduces two new filter functions to replace the original projector-based filter:

Immanant Filter: Uses the immanant $\text{Imm}_\mu(U)$ as the filter function. This reduces the number of algebraic terms compared to the original permanent-based approach and eliminates the need for CG coefficients.
Character Filter: Uses the character $\chi_\mu(U)$ $χ_{μ} (U)$ (the trace of the representation) as the filter function.
- Advantages: Characters are computationally efficient (polynomial time evaluation via Schur polynomials) and possess a constant variance (equal to 1) regardless of the irrep or group size, due to the orthogonality of characters.

C. Experimental Relaxation

The paper argues that the filtering process remains valid even if the input state is not a pure Fock state but a weak coherent state (which contains a superposition of photon numbers, including vacuum and multi-photon sectors).

By using intensity measurements (e.g., single-photon avalanche diode "click" detection) instead of PNR detectors, the scheme can estimate the fidelity of the single-photon subspace even in the presence of photon loss and gain.
Theoretical analysis shows that the character filter isolates the desired parameter ( $p_\mu$ ) even when the system evolves in an extended Hilbert space containing loss/gain sectors.

3. Key Contributions

Elimination of CG Coefficients and Permanents: The proposed filters (Immanant and Character) remove the need to compute Clebsch–Gordan coefficients and matrix permanents, which were the primary computational bottlenecks of the original scheme.
Kostant Relation Application: The paper successfully applies Kostant's relation to RB, demonstrating that zero-weight states can serve as effective filters to extract decay parameters.
Character Filter Efficiency: The introduction of the character filter provides a method with polynomial computational complexity and constant variance, making it the most scalable option for large mode numbers ( $m$ ) and many irreps.
Experimental Feasibility: The work demonstrates that the protocol can be implemented using weak coherent states and intensity measurements, significantly lowering the barrier to entry for experimentalists compared to Fock state preparation and PNR detection.

4. Results

Numerical Validation: The author performed numerical simulations on a two-mode system ( $m=2$ $m = 2$ ) allowing for 0, 1, and 2 photons (simulating loss and gain).
- The scheme successfully estimated the average gate fidelity (AGF) of the single-photon subspace.
- For fidelities above 0.95, the estimation error was within 2%.
- The method slightly overestimates fidelity due to contributions from additional irreps in the extended Hilbert space, but the trend remains accurate.
Variance Analysis: The character filter was shown to have a variance of 1, independent of the system size, whereas immanants and permanents have variances that can grow with system complexity.
Comparison:
- Original Scheme: High algebraic overhead, requires CG coefficients, exponential scaling with $n$ (photon number) for permanents.
- Immanant Filter: Reduces algebraic terms and removes CGs, but evaluation cost can still be high if immanants are complex. Best when $n \approx m$ .
- Character Filter: Most efficient. Polynomial evaluation cost, constant variance, no permanents or immanants required. Preferred for large-scale post-processing.

5. Significance

This work significantly advances the practical application of Randomized Benchmarking in continuous-variable (bosonic) quantum computing.

Scalability: By replacing #P-hard permanent calculations with polynomial-time character evaluations, the method becomes scalable to larger interferometers and higher photon numbers.
Accessibility: The relaxation of experimental requirements (allowing weak coherent states and simple intensity detectors) makes bosonic RB accessible to a wider range of quantum optics laboratories that lack advanced Fock state sources or PNR detectors.
Robustness: The scheme maintains the core advantages of RB (SPAM robustness) while providing a clearer theoretical path for data analysis, potentially accelerating the characterization of photonic quantum processors.

In conclusion, the paper provides a rigorous theoretical foundation and numerical evidence that character-based filtering is the optimal path forward for benchmarking passive bosonic devices, offering a balance of computational efficiency, statistical stability, and experimental feasibility.