Year: 2025
Type: conference-paper
Abstract: The AlphaZero framework provides a standard way of combining Monte Carlo planning with prior knowledge provided by a previously trained policy-value neural network. AlphaZero usually assumes that the environment on which the neural network was trained will not change at test time, which constrains its applicability. In this paper, we analyze the problem of deploying AlphaZero agents in potentially changed test environments and demonstrate how the combination of simple modifications to the standard framework can significantly boost performance, even in settings with a low planning budget available. The code is publicly available on GitHub.
Source: Lirias
Authors: Tamassia, Isidoro, Böhmer, Wendelin
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7080122099
FWCI:
Cites:
Cited by:
Related to: