gymCICY

Heterotic line bundle models are the most succesful construction of realistic compactification for the heterotic string. The number of configurations to consider in systematic scans scales exponentially with \(h^{(1,1)}\). Hence, for larger values smart ways are needed to parse the landscape of possible configurations.

The Package

gymCICY lets you explore heterotic line bundle models via the successfull OpenAI gym interface. This python package contains two different settings to construct sums of line bundles. First, the stacking environment, which simulates the systematic construction of line bundle sums via stacking of five line bundles. Second, in flipping agents interact with the charges of the line bundles by either increasing or decreasing them with \(\pm 1\). You can install the package with

pip install --user git+https://github.com/robin-schneider/gymCICY

it uses the pyCICY package for the reward-computations.