Motivated by the recently-established connection between Jarzynski’s equality and the theoretical framework of Stochastic Normalizing Flows, we investigate a protocol relying on out-of-equilibrium lattice Monte Carlo simulations to mitigate the infamous computational problem of topological freezing. We test our proposal on 2d CP^N−1 models and compare our results with those obtained adopting the Parallel Tempering on Boundary Conditions proposed by M. Hasenbusch, obtaining comparable performances. Our work thus sets the stage for future applications combining our Monte Carlo setup with machine learning techniques.

中文翻译：

---

使用非平衡模拟减轻拓扑冻结

受最近在 Jarzynski 等式与随机归一化流理论框架之间建立的联系的启发，我们研究了一种依赖于非平衡晶格蒙特卡罗模拟的协议，以缓解臭名昭著的拓扑冻结计算问题。我们在 2 d CP ^{N −1}模型上测试了我们的建议，并将我们的结果与采用 M. Hasenbusch 提出的边界条件平行回火所获得的结果进行比较，获得了可比较的性能。因此，我们的工作为未来将蒙特卡罗设置与机器学习技术相结合的应用奠定了基础。