Piqasso tackles Cat Qubits

Inspiration

We're a band of brothers who loves physics, loves quantum computers, and like everyone participating in this challenge have come to gain a deeper understanding and appreciation of cat qubits.

What it does

Our PPO-inspired surrogate neural network framework characterizes the behavior of states in the Fock space (such as quantum harmonic oscillators and cat qubits). We applied our model to observe these states under specific drift conditions, during single-qubit gates, and during the creation of mooncat qubits.

How we built it

We first examined the challenge Jupyter notebook and created an end-to-end CMAS-ES optimization algorithm. We then made a PPO optimization algorithm and compared the outputted $\epsilon_d$ and $g_2$ control parameters with our CMAS-ES model. Finally, we decided to create a PPO-inspired surrogate neural network to reduce the number of differential equations we needed to solve, thereby expediting our simulation process.

Challenges we ran into

Estimating Observables
Modeling Drift
Stabilizing Optimization
Understanding Dissipative Quantum Computing Regimes

Accomplishments that we're proud of

Running our model on Google Colab to use H100 GPUs
Gaining a deeper physical understanding of the cat qubit system through simulational physics
Building three functional optimization algorithms and benchmarking them to discover the best one

What we learned

The fantastical world of cat qubits
How to effectively benchmark simulations and determine the best optimization model

What's next for Piqasso tackles Cat Qubits

We hope to learn more about quantum computing in the near future and get involved with quantum computing research: bringing this exotic technology to the world one qubit at a time!