Dynamic Validation Instances

该数据集包含动态生成的不同大小的验证实例，旨在评估针对每个领域的一般化动作策略的扩展行为。在生成实例的过程中，数据集会持续生成直到达到对覆盖性能的特定置信度，以确保对策略性能的全面评估。该数据集涵盖了9个领域，用于实验研究，其任务是对动作策略的扩展行为进行评估。

This dataset contains dynamically generated validation instances of varying sizes, aimed at evaluating the scaling behavior of generalized action policies across each individual domain. During the instance generation process, the dataset continues generating instances until a specified confidence level in coverage performance is achieved, ensuring a comprehensive assessment of the policy's performance. This dataset encompasses 9 domains for experimental research, with its core task being to evaluate the scaling behavior of action policies.