Datasets:

TheFinAI
/

gr-QA

pretty_name: Plutus · GRFinQA (FinBen task format)
license: apache-2.0
language:
  - el
task_categories:
  - question-answering
size_categories:
  - n<1K
tags:
  - finance
  - greek
  - benchmark
  - thefinai
  - plutus
  - finben
extra_gated_heading: Request access to Plutus · GRFinQA (FinBen task format)
extra_gated_description: >-
  Released by The Fin AI for research. Access is granted automatically after you
  complete this short form.
extra_gated_button_content: Agree and access
extra_gated_prompt: >-
  By accessing this dataset you agree to its license and to cite the Plutus
  paper in any resulting publication.
extra_gated_fields:
  Full name: text
  Affiliation: text
  Country: country
  Intended use:
    type: select
    options:
      - Research
      - Education
      - Commercial evaluation
      - Other
  I agree to the terms above and will cite the paper: checkbox

Plutus · GRFinQA (FinBen task format)

📄 Paper · 🤗 Collection · 💻 FinBen task · 🌐 The Fin AI

Part of Plutus — Plutus: Benchmarking Large Language Models in Low-Resource Greek Finance (arXiv:2502.18772).

Task: question answering (multiple choice, Greek finance exam questions).

This repository is the copy loaded by the FinBen evaluation harness (tasks/plutus/grQA.yaml). It contains the same examples as plutus-QA, which has the full dataset card; only the prompt/answer formatting differs slightly.

Quick Start

from datasets import load_dataset

ds = load_dataset("TheFinAI/gr-QA", split="test")
print(ds[0])

Split	Rows
`train`	267
`validation`	48
`test`	225

License

Same license as plutus-QA (Apache 2.0), which contains the same questions.

Citation

@misc{peng2025plutusbenchmarkinglargelanguage,
      title={Plutus: Benchmarking Large Language Models in Low-Resource Greek Finance}, 
      author={Xueqing Peng and Triantafillos Papadopoulos and Efstathia Soufleri and Polydoros Giannouris and Ruoyu Xiang and Yan Wang and Lingfei Qian and Jimin Huang and Qianqian Xie and Sophia Ananiadou},
      year={2025},
      eprint={2502.18772},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2502.18772}, 
}