⚛️ quantum physics

Multimode rotationally symmetric bosonic codes from group-theoretic construction

This paper introduces a new family of multi-mode, rotationally symmetric bosonic codes derived from a group-theoretic framework that enables exact correction of correlated dephasing and full Pauli group implementation via linear optics, while uniquely eliminating the trade-off between dephasing protection and photon loss found in single-mode analogues.

Original authors: Rabsan Galib Ahmed, Adithi Udupa, Giulia Ferrini

Published 2026-03-26

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Rabsan Galib Ahmed, Adithi Udupa, Giulia Ferrini

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 are trying to send a fragile message across a stormy ocean. In the world of quantum computing, this "message" is a piece of information (a qubit), and the "ocean" is the physical hardware, which is constantly buffeted by waves of noise. Two of the biggest waves are photon loss (the message getting dropped or lost) and dephasing (the message getting scrambled or confused).

For a long time, scientists have been trying to build "lifeboats" (error-correcting codes) to keep these messages safe. Most lifeboats are built for a single passenger (a single mode of light). But this new paper introduces a revolutionary new design: a multi-passenger, rotationally symmetric lifeboat that is much harder to sink.

Here is the breakdown of what the researchers did, using simple analogies:

1. The Old Way: Building the Boat First, Then the Oars

Traditionally, scientists would design a specific boat (a code) to hold their message, and then try to figure out how to steer it (apply logic gates) using whatever tools they had. It was like building a car and then realizing you need a special, complicated key to start the engine.

The New Approach: The authors flipped the script. They said, "Let's decide exactly how we want to steer the car first (using simple, easy tools like linear optics, which are just mirrors and beam-splitters), and then design the car to fit those steering wheels."

The Analogy: Instead of building a car and then trying to invent a steering wheel, they designed a car specifically shaped so that a simple, standard steering wheel fits perfectly.

2. The "Rotational Symmetry" Secret Sauce

The core of their new code is Rotational Symmetry.
Imagine a pizza cut into slices. If you rotate the pizza by one slice, it looks exactly the same.

Single-mode codes (Old): These are like a single pizza. If you lose a slice (photon loss), you can tell something is wrong. But if the pizza gets "scrambled" (dephasing), it's hard to fix. Also, there's a trade-off: making the pizza cut into more slices (higher symmetry) helps with losing slices but makes it harder to fix scrambling.
Multi-mode codes (New): The authors created a two-pizza system. They link two pizzas together.
- The Magic: By linking them, they can rotate the pizzas in a way that if one gets scrambled, the other helps fix it.
- The Result: They broke the "trade-off rule." In the old single-pizza world, you had to choose between protecting against lost slices or scrambled toppings. In this new two-pizza world, you get protection against both at the same time.

3. The "Binomial" Lifeboat

They tested a specific version of this called the Dual-Rail Binomial Code.

The Analogy: Think of a "Binomial" code as a specific pattern of how you stack your cargo. It's not just a random pile; it's a carefully balanced stack.
The "Dual-Rail": Instead of putting all the cargo in one truck (one mode), they split it between two trucks (two modes) that are linked by a magical bridge (a beam-splitter).
The Outcome: When they ran simulations, this two-truck system performed significantly better than the single-truck system, especially against the "scrambling" noise (dephasing). It was like finding a lifeboat that doesn't just float better; it actively corrects its own tilt.

4. The "Ghost" Noise (Correlated Dephasing)

Sometimes, the noise isn't random; it's coordinated. Imagine a storm that hits both of your trucks at the exact same time in the exact same way. This is called correlated dephasing.

The Problem: Usually, if both trucks get hit the same way, you can't tell if they are damaged or just moving together.
The Solution: The authors found a clever trick using a "control truck" (an auxiliary qubit). By performing a specific swap operation (like a magic dance move), they can move the "clean" version of the message from the damaged trucks to the control truck, effectively erasing the noise.
The Result: They proved mathematically that this new code can perfectly correct this specific type of coordinated noise. It's like having a backup generator that instantly kicks in the moment the main power grid fails, without any delay.

5. Why This Matters

Efficiency: Quantum computers need thousands of physical parts to make one stable "logical" part. This new code reduces the number of parts needed because it is so good at fixing errors on its own.
Simplicity: The "steering wheels" (logic gates) for this code are simple optical tools (mirrors and beam splitters) that are easy to build in a lab, rather than complex, hard-to-control pulses.
Scalability: Because it works so well against the two biggest enemies of quantum computers (loss and scrambling), it brings us closer to building a large-scale, fault-tolerant quantum computer that can actually solve real-world problems.

In a Nutshell

The researchers invented a new way to store quantum information by using two linked modes of light instead of one. By designing the code around the simple tools we already have, they created a system that is stronger, more efficient, and capable of fixing two types of errors simultaneously—something previous designs couldn't do. It's a major step toward making quantum computers reliable enough to change the world.

1. Problem Statement

Bosonic codes encode quantum information into the infinite-dimensional Hilbert space of harmonic oscillators (e.g., microwave cavities) to protect against dominant noise sources like photon loss and dephasing. While single-mode rotationally symmetric bosonic (RSB) codes (e.g., binomial or cat codes) offer protection, they face a fundamental trade-off: increasing the order of rotational symmetry ( $N$ ) improves protection against photon loss but degrades protection against dephasing. Furthermore, implementing logical gates (especially the $X$ -gate) on single-mode RSB codes often requires complex, optimized control pulses or auxiliary qubits.

Existing group-theoretic constructions (specifically Ref. [14]) inverted the design paradigm by selecting a group of logical gates (e.g., the Pauli group) to be implemented by simple physical operations (e.g., linear optics) and deriving the code from this requirement. However, these constructions did not fully exploit general symmetries to enhance error correction properties, nor did they address the loss-dephasing trade-off in multi-mode settings.

2. Methodology

The authors propose a new family of Multi-mode Rotationally Symmetric Bosonic (RSB) codes by extending the group-theoretic framework of Ref. [14] with two critical innovations:

Group-Theoretic Construction with Mode Mixing:
- Logical Group: The logical Pauli group for a qudit of dimension $d$ is generated by $X$ and $Z$ .
- Physical Representation: Instead of mapping logical gates directly to single-mode operations, the authors introduce a passive linear optical network (a beam splitter operation, $\hat{U}_{BS}$ ) that mixes $d$ physical bosonic modes.
- Mapping:
  - Logical $Z$ is implemented via a phase rotation on one mode, conjugated by the beam splitter: $\hat{U}_{BS} e^{i \frac{2\pi}{Nd} \hat{a}_1^\dagger \hat{a}_1} \hat{U}_{BS}^\dagger$ .
  - Logical $X$ is implemented via a SWAP-like operation (using beam splitters) conjugated by the beam splitter.
- Codewords: The resulting codewords are superpositions of Fock states where the photon numbers in each mode follow specific modular arithmetic rules (multiples of $N$ ), entangled via the beam splitter. For a qubit ( $d=2$ ), the codewords are:
  $|0_N\rangle = \hat{U}_{BS} \sum_{m,n} f_{mn} |2mN, (2n+1)N\rangle$
  $|1_N\rangle = \hat{U}_{BS} \sum_{m,n} f_{mn} |(2n+1)N, 2mN\rangle$
Specific Instance: Dual-Rail Binomial Code (DRBC):
The authors focus on a specific instance where the coefficients $f_{mn}$ are chosen to create a "dual-rail binomial" structure (a generalization of the single-mode binomial code). They analyze the performance of this code under loss and dephasing channels.
Error Correction Analysis:
- Knill-Laflamme (KL) Conditions: The authors analytically derive the conditions on the coefficients $f_{mn}$ and beam-splitter angles ( $\delta, \phi$ ) required to satisfy the KL conditions up to the first order of noise strength.
- Recovery Maps: They propose recovery procedures involving modular number measurements (generalized parity measurements) followed by unitary corrections.
- Correlated Dephasing: They design a specific quantum circuit using controlled-X (CX) gates to exactly correct correlated dephasing noise (noise affecting the total photon number $\hat{n}_1 + \hat{n}_2$ ).

3. Key Contributions

Resolution of the Loss-Dephasing Trade-off:
Unlike single-mode RSB codes, where increasing $N$ improves loss protection but worsens dephasing protection, the two-mode DRBC exhibits a simultaneous improvement in both error correction capabilities as $N$ increases (up to an optimal point). This breaks the traditional trade-off.
Exact Correction of Correlated Dephasing:
The authors demonstrate that any two-mode RSB code in their family can exactly correct correlated dephasing noise (common in flux-tunable couplers) using a specific circuit involving controlled operations and an auxiliary mode. This is a significant advantage over single-mode codes which cannot correct this specific noise type exactly.
Linear-Optics Implementation of Logical Gates:
The logical Pauli group ( $X, Z$ ) is implemented entirely using passive linear optics (beam splitters and phase shifters), avoiding the need for complex, non-linear control pulses required by single-mode RSB codes.
Qudit Encoding:
The construction naturally extends to encoding qudits of arbitrary dimension $d$ using $d$ bosonic modes, offering a scalable path for high-dimensional quantum information processing.

4. Results

Numerical Benchmarks:
- Dephasing: For the DRBC with $K=2$ (binomial order), numerical simulations show that the logical fidelity under dephasing noise significantly outperforms both the single-mode binomial code and the two-mode CLY code. The performance improves as $N$ increases (e.g., $N=2, 4, 6, 8$ ), with no observed degradation.
- Loss: The code performs comparably to the single-mode binomial code under pure loss, maintaining the "break-even" regime (outperforming the trivial encoding).
- Combined Noise: Under combined loss and dephasing, the two-mode code shows superior overall performance compared to single-mode analogues.
Analytical Insights:
- KL Conditions: For dephasing, the KL conditions are satisfied up to first order when the beam-splitter angle $\delta = \pi/4$ (or $3\pi/4$ ) and $\phi = \pi/(2N)$ .
- Distinguishability: The authors explain the improved dephasing performance via phase distinguishability. In single-mode codes, phase distributions of logical states overlap on a circle ( $S^1$ ). In two-mode codes, the phase space is a torus ( $S^1 \times S^1$ ). By tuning encoding angles, the probability distributions of the dual codewords can be made mutually exclusive on the torus, preventing overlap even under large rotation errors.
Correlated Dephasing:
The proposed circuit successfully swaps the noise-free state from the data rail to an auxiliary rail, effectively removing the correlated phase noise without disturbing the logical information.

5. Significance

This work represents a paradigm shift in bosonic code design by successfully integrating group-theoretic constraints with multi-mode entanglement to overcome fundamental limitations of single-mode codes.

Scalability: By resolving the loss-dephasing trade-off and enabling exact correction of correlated noise, these codes reduce the overhead required for fault-tolerant quantum computing.
Hardware Compatibility: The reliance on passive linear optics for logical gates makes these codes highly compatible with existing superconducting circuit architectures (e.g., microwave cavities with tunable couplers).
Robustness: The ability to handle non-Markovian dephasing (random telegraph noise) and correlated noise makes these codes particularly suitable for near-term quantum hardware where such noise sources are prevalent.

In summary, the paper introduces a robust, multi-mode bosonic code family that leverages group theory and linear optics to achieve superior error correction performance, specifically breaking the trade-off between loss and dephasing resilience that has hindered previous single-mode approaches.