BPS spectroscopy with reinforcement learning

Recent advancements in quantum computing (QC) and machine learning (ML) have sparked considerable interest in the integration of these two cutting-edge fields. Among the various ML techniques, reinforcement learning (RL) stands out for its ability to address complex sequential decision-making problems. RL has already demonstrated substantial success in the classical ML community. Now, the emerging field of Quantum Reinforcement Learning (QRL) seeks to enhance RL algorithms by...

Reinforcement Learning (RL) is a rapidly growing area of machine learning that finds its application in a broad range of domains, from finance and healthcare to robotics and gaming. Compared to other machine learning techniques, RL agents learn from their own experiences using trial and error, and improve their performance over time. However, assessing RL models can be challenging, which makes it difficult to interpret their behaviour. While reward is a widely used metric to ...

BPS quivers for N=2 SU(N) gauge theories are derived via geometric engineering from derived categories of toric Calabi-Yau threefolds. While the outcome is in agreement of previous low energy constructions, the geometric approach leads to several new results. An absence of walls conjecture is formulated for all values of N, relating the field theory BPS spectrum to large radius D-brane bound states. Supporting evidence is presented as explicit computations of BPS degeneracies...

We study the BPS spectrum of four-dimensional $\mathcal{N}=2$ supersymmetric Yang-Mills theory with gauge group $SU(2)$ and a massive adjoint hypermultiplet, which has an extremely intricate structure with infinite spectrum in all chambers of its Coulomb moduli space, and is not well understood. We build on previous results by employing the BPS quiver description of the spectrum, and explore the qualitative features in detail using numerical techniques. We find novel and unex...

Reinforcement learning studies how an agent should interact with an environment to maximize its cumulative reward. A standard way to study this question abstractly is to ask how many samples an agent needs from the environment to learn an optimal policy for a $\gamma$-discounted Markov decision process (MDP). For such an MDP, we design quantum algorithms that approximate an optimal policy ($\pi^*$), the optimal value function ($v^*$), and the optimal $Q$-function ($q^*$), ass...

Efficient numerical optimization methods can improve performance and reduce the environmental impact of computing in many applications. This work presents a proof-of-concept study combining primitive state representations and agent-environment interactions as first-order optimizers in the setting of budget-limited optimization. Through reinforcement learning (RL) over a set of training instances of an optimization problem class, optimal policies for sequential update selectio...

This monograph is aiming to serve as a primary introduction to spectral networks, by presenting them and their many applications from the scope of geometry and mathematical physics in a unified way. We do not attempt to treat these two approaches separately but rather provide broad motivation and the necessary background to reach to the frontiers of this fast-evolving field of research, explaining simultaneously the relevant geometric and physical aspects. The targeted audien...

A large class of N=2 quantum field theories admits a BPS quiver description and the study of their BPS spectra is then reduced to a representation theory problem. In such theories the coupling to a line defect can be modelled by framed quivers. The associated spectral problem characterises the line defect completely. Framed BPS states can be thought of as BPS particles bound to the defect. We identify the framed BPS degeneracies with certain enumerative invariants associated ...

We exactly evaluate the partition function (index) of N=4 supersymmetric quiver quantum mechanics in the Higgs phase by using the localization techniques. We show that the path integral is localized at the fixed points, which are obtained by solving the BRST equations, and D-term and F-term conditions. We turn on background gauge fields of R-symmetries for the chiral multiplets corresponding to the arrows between quiver nodes, but the partition function does not depend on the...

Extending results of arXiv:1112.3984, we show that all rank 1 N=2 SCFT's in the sequence H_1, H_2, D_4 E_6, E_7, E_8 have canonical finite BPS chambers containing precisely 2 h(F)=12(Delta-1) hypermultiplets. The BPS spectrum of the canonical BPS chambers saturates the conformal central charge c, and satisfies some intriguing numerology.