README.md

---
datasets:
- idavidrein/strategyqa
language:
- en
---

# StrategyQA - Processed Subset

This dataset is a curated subset of **StrategyQA**, organized into nested tiers for progressive benchmarking and scalable oversight experiments.

## Tier Structure
The samples are organized into three cumulative tiers:
* **Small**: The base subset.
* **Medium**: Includes all "Small" samples plus additional entries.
* **Large**: Includes all "Medium" samples, representing the full subset.

## Citation
If you use this dataset, please cite the original work:

```bibtex
@article{Geva2021DidAU,
  title={Did Aristotle Use a Laptop? A Question Answering Benchmark with Implicit Reasoning Strategies},
  author={Mor Geva and Daniel Khashabi and Elad Segal and Tushar Khot and Dan Roth and Jonathan Berant},
  journal={Transactions of the Association for Computational Linguistics},
  year={2021},
  volume={9},
  pages={346-361},
  url={[https://api.semanticscholar.org/CorpusID:230799347](https://api.semanticscholar.org/CorpusID:230799347)}
}