Benchmarking Results
收藏arXiv2025-09-30 收录
下载链接:
https://github.com/HumanCompatibleAI/imitation
下载链接
链接失效反馈官方服务:
资源简介:
该数据集汇总了一张表格,其中详细记录了在进行基准测试期间获得的结果,包括置信区间。此外,在基准测试文件夹中,还包含了用于训练算法的超参数信息。这项任务旨在对模仿和学习奖励算法进行基准测试。
This dataset compiles a table that comprehensively records the results obtained during benchmarking, including confidence intervals. Additionally, the benchmarking folder contains the hyperparameter information used for training the algorithms. This task aims to benchmark imitation and reward learning algorithms.
提供机构:
HumanCompatibleAI



