Skip to content

Dataset · hails

LogiQA logic tasks from the AGIEval benchmark

hails/agieval-logiqa-en

English subset of LogiQA logic tasks from Microsoft's AGIEval benchmark for evaluating LLM reasoning.

What's inside

It contains logical reasoning task texts from the English-language part of LogiQA, which is part of the AGIEval benchmark. The dataset contains less than a thousand entries in English, taken from Microsoft's official GitHub repository and processed similarly to other AGIEval sets.

What you can do with it

  • Evaluating the logical reasoning capabilities of language models
  • Comparing LLMs on the standard AGIEval benchmark
Languages
English
Size
up to 1K rows
Downloads this month
20K
Likes
0

How to load it

from datasets import load_dataset

ds = load_dataset("hails/agieval-logiqa-en")