BootSTOP (Bootstrap STochastic OPtimiser)

Contents

Overview

BootSTOP is a Python package for determining CFT data (OPE-coefficients squared and scaling dimensions) which minimise a theory’s truncated crossing equation. To do this the code can apply either a custom PyTorch implementation of the Soft-Actor-Critic algorithm or one of the algorithms within the PyGMO package (information about PyGMO can be found on the PyGMO website).

At present the crossing equation for each of the following CFTs is coded within BootSTOP: 1D defect CFT (see [4]), 2D c=1 compactified boson CFT (see [1,2]) and the 6D (2,0) SCFT (see [3]). See [1-4] and the references within for comprehensive background.

Installation

For installation instructions see our getting started page.

Then you will need to follow download the blocks from our block page from where you should copy the pre-generated conformal block files and place them within the relevant block_lattices folder in your local BootSTOP repository.

Running the code

This repository contains several configured config and run files so that you can immediately start solving. For example, executing the command:

   python3 run_pygmo_1d_deriv.py

will apply PyGMO’s IPOPT algorithm to the (derivative form) crossing equation of the 1D defect CFT truncated to 62 operators.

References

[1] G. Kantor, V. Niarchos and C. Papageorgakis, Solving Conformal Field Theories with Artificial Intelligence, Phys. Rev. Lett. 128 (2022) arXiv:2108.08859
[2] G. Kantor, V. Niarchos and C. Papageorgakis, Conformal bootstrap with reinforcement learning, Phys. Rev. D 105 (2022) arXiv:2108.09330
[3] G. Kantor, V. Niarchos, C. Papageorgakis and P. Richmond, 6D (2,0) bootstrap with soft-actor-critic algorithm, Phys. Rev. D 107 (2023) arXiv:2209.02801
[4] V. Niarchos, C. Papageorgakis, P. Richmond, A. G. Stapleton and M. Wooley, Bootstrability in Line-Defect CFT with Improved Truncation Methods, arXiv:2306.15730