Spico/TaskLAMA

--- license: cc-by-4.0 --- # TaskLAMA This is an unofficial upload of the [TaskLAMA](https://arxiv.org/abs/2308.15299) data. TaskLAMA is a novel dataset for Structured Complex Task Decomposition (SCTD). Some of the data statistics could be found at [Spico197/TaskLAMA](https://github.com/Spico197/TaskLAMA) . ## Citation ```bibtex @misc{yuan2023tasklama, title={TaskLAMA: Probing the Complex Task Understanding of Language Models}, author={Quan Yuan and Mehran Kazemi and Xin Xu and Isaac Noble and Vaiva Imbrasaite and Deepak Ramachandran}, year={2023}, eprint={2308.15299}, archivePrefix={arXiv}, primaryClass={cs.CL} } ```--- 许可证：CC-BY-4.0 --- # TaskLAMA 本仓库为[TaskLAMA](https://arxiv.org/abs/2308.15299)论文对应数据集的非官方上传版本。 TaskLAMA是面向结构化复杂任务分解（Structured Complex Task Decomposition, SCTD）的新型数据集。 部分数据集统计信息可参见[Spico197/TaskLAMA](https://github.com/Spico197/TaskLAMA)。 ## 引用 bibtex @misc{yuan2023tasklama, title={TaskLAMA: Probing the Complex Task Understanding of Language Models}, author={Quan Yuan and Mehran Kazemi and Xin Xu and Isaac Noble and Vaiva Imbrasaite and Deepak Ramachandran}, year={2023}, eprint={2308.15299}, archivePrefix={arXiv}, primaryClass={cs.CL} }

原始信息汇总

TaskLAMA

TaskLAMA是一个用于结构化复杂任务分解（SCTD）的新型数据集。

引用

bibtex @misc{yuan2023tasklama, title={TaskLAMA: Probing the Complex Task Understanding of Language Models}, author={Quan Yuan and Mehran Kazemi and Xin Xu and Isaac Noble and Vaiva Imbrasaite and Deepak Ramachandran}, year={2023}, eprint={2308.15299}, archivePrefix={arXiv}, primaryClass={cs.CL} }

搜集汇总

背景与挑战

背景概述

TaskLAMA是一个专注于结构化复杂任务分解（SCTD）的数据集，包含1,611行数据，提供了任务的分解步骤、假设和依赖关系。该数据集适用于测试语言模型对复杂任务的理解能力，相关研究发表在arXiv上。