Evolution Simulation Dataset
收藏arXiv2025-09-30 收录
下载链接：
https://github.com/oist/emevo
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含了一个基于人群的分布式进化模拟框架的结果，该框架模拟了基于环境条件的生物合理奖励功能的进化。此外，该数据集捕捉了在不同环境条件，如食物密度和动作奖励下，奖励功能的进化动态。模拟范围涉及100至200个智能体。该数据集的任务是基于食物可用性和动作模拟智能体奖励的进化。
This dataset encompasses results derived from a population-based distributed evolutionary simulation framework. This framework models the evolution of biologically plausible reward functions conditioned on environmental parameters. Additionally, the dataset captures the evolutionary dynamics of reward functions under varied environmental conditions, such as food density and action rewards. The simulations involve populations of 100 to 200 agents. The core task of this dataset is to simulate the evolution of agent rewards based on food availability and action-related factors.
提供机构：
Authors of the paper
