Ebisu — Japanese financial evaluation
Collection
Ebisu (arXiv:2602.01479): JF-ICR and JF-TE. • 4 items • Updated
Request access to Ebisu JF-TE
Ebisu JF-TE is released by The Fin AI for research. Access is granted automatically after you complete this short form.
By accessing this dataset you agree to its license and to cite the Ebisu paper in any resulting publication.
Log in or Sign Up to review the conditions and access this dataset content.
Ebisu JF-TE
📄 Paper · 💻 Code · 🌐 The Fin AI
Part of Ebisu — Ebisu: Benchmarking Large Language Models in Japanese Finance (arXiv:2602.01479). Formerly
TheFinAI/JF-TE(the old name redirects here).
JF-TE (Japanese Financial Term Extraction) asks a model to extract financial terms — including nested nominal compounds — from Japanese corporate disclosures. The paper evaluates it with Maximal Financial Term F1 and HitRate@K for nested terms.
This public release contains 58 examples (train split, single jf_te.jsonl file).
Quick Start
from datasets import load_dataset
ds = load_dataset("TheFinAI/jp-te", split="train")
print(ds[0]["query"])
print(ds[0]["answer"])
| Field | Description |
|---|---|
query |
Instruction prompt with the Japanese passage |
answer |
Gold financial terms (nested lists) |
id |
Example id |
Citation
@misc{peng2026ebisubenchmarkinglargelanguage,
title={Ebisu: Benchmarking Large Language Models in Japanese Finance},
author={Xueqing Peng and Ruoyu Xiang and Fan Zhang and Mingzi Song and Mingyang Jiang and Yan Wang and Lingfei Qian and Taiki Hara and Yuqing Guo and Jimin Huang and Junichi Tsujii and Sophia Ananiadou},
year={2026},
eprint={2602.01479},
archivePrefix={arXiv},
primaryClass={cs.CL},
url={https://arxiv.org/abs/2602.01479},
}
- Downloads last month
- 13