Synthetic MAX-SAT Models
收藏arXiv2025-09-30 收录
下载链接:
https://github.com/mohitKULeuven/HassleWithLocalSearch
下载链接
链接失效反馈官方服务:
资源简介:
该数据集包含通过枚举文字的析取以及随机权重下的硬约束和软约束生成的合成MAX-SAT模型。此外,为每个任务生成了五个随机模型,这些模型使用了不同的随机化种子。数据集的规模是可变的,针对n=8、10、12、15的情况生成了不同约束条件的模型。任务内容是从上下文示例中学习MAX-SAT模型。
This dataset comprises synthetic MAX-SAT models generated using enumerated disjunctions of literals, paired with hard and soft constraints under random weights. Furthermore, five random models were generated for each task, each employing a distinct randomization seed. The dataset is scalable in size, with models featuring varying constraint configurations generated for parameter settings of n=8, 10, 12, and 15. The task objective is to learn MAX-SAT models from in-context examples.
提供机构:
Authors of the paper



