TaskBench

What is TaskBench?

TaskBench aims to evaluate the capability of LLMs in task automation, containing 28,271 samples spanning 3 critical stages: task decomposition, tool…

Released

2023-11-30

Evaluates

智能体, Agent, NeurIPS 2024, 任务执行, Task Execution, 不支持, Unsupported

Openness

unknown

Importer

OpenCompass Hub

Reported scores

0

Source provenance

• Original evidence https://hub.opencompass.org.cn/dataset-detail/TaskBench

TaskBench paper and code

• Paper https://arxiv.org/pdf/2311.18760
• Code https://github.com/microsoft/JARVIS

Which sources cite TaskBench?

• TaskBench

Related 智能体, Agent, NeurIPS 2024, 任务执行, Task Execution, 不支持, Unsupported benchmarks

• A-Bench
• A-OKVQA
• AbsPyramid
• ActionAtlas
• Ada-LEval
• AECBench
• AFQMC
• AgentBoard