Ziying Jia | Searcharxiv

arXiv2025

Superconductivity in imbalanced bilayer Hubbard model: enhanced $d$-wave and weakened $s^\pm$-wave pairing

We investigate the bilayer model with two layers of imbalanced densities coupled by the interlayer hybridization. Using the large-scale dynamical cluster quantum Monte Carlo simulation, we discovered that increased hybridization induces a transition in the superconducting pairing from $d$-wave to $s^{\pm}$-wave and the superconducting $T_c$ of $d$-wave pairing exhibits a non-monotonic dependence on the density imbalance. Remarkably, the optimal superconductivity(SC) occurs at a moderate imbalance. Our results support the possibility of $T_c$ enhancement in composite picture where the underdoped layer provides the pairing strength while the overdoped layer promotes the phase coherence. In addition, the SC can be possibly hosted by a single layer, which is reminiscent of our recent exploration on the trilayer Hubbard model. Our present study thus provides new insight that the SC can be enhanced via the layer differentiation.

cond-mat.supr-con↗

arXiv2025

Pessimism Principle Can Be Effective: Towards a Framework for Zero-Shot Transfer Reinforcement Learning

Transfer reinforcement learning aims to derive a near-optimal policy for a target environment with limited data by leveraging abundant data from related source domains. However, it faces two key challenges: the lack of performance guarantees for the transferred policy, which can lead to undesired actions, and the risk of negative transfer when multiple source domains are involved. We propose a novel framework based on the pessimism principle, which constructs and optimizes a conservative estimation of the target domain's performance. Our framework effectively addresses the two challenges by providing an optimized lower bound on target performance, ensuring safe and reliable decisions, and by exhibiting monotonic improvement with respect to the quality of the source domains, thereby avoiding negative transfer. We construct two types of conservative estimations, rigorously characterize their effectiveness, and develop efficient distributed algorithms with convergence guarantees. Our framework provides a theoretically sound and practically robust solution for transfer learning in reinforcement learning.

cs.LG↗