datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
legalbench
Dataset Card for Dataset Name
Homepage: https://hazyresearch.stanford.edu/legalbench/
Repository: https://github.com/HazyResearch/legalbench/
Paper: https://arxiv.org/abs/2308.11462
Dataset Description
Dataset Summary
The LegalBench project is an ongoing open science effort to collaboratively curate tasks for evaluating legal reasoning in English large language models (LLMs). The benchmark currently consists of 162 tasks gathered from 40… See the full description on the dataset page: https://huggingface.co/datasets/nguha/legalbench.takedown-notices
Takedown notices received by the Hugging Face team
Please click on Files and versions to browse them
Also check out our:
Terms of Service
Community Code of Conduct
Content Guidelines
legalbench-staging
Dataset Card for Dataset Name
Homepage: https://hazyresearch.stanford.edu/legalbench/
Repository: https://github.com/HazyResearch/legalbench/
Paper: https://arxiv.org/abs/2308.11462
Dataset Description
Dataset Summary
The LegalBench project is an ongoing open science effort to collaboratively curate tasks for evaluating legal reasoning in English large language models (LLMs). The benchmark currently consists of 162 tasks gathered from 40… See the full description on the dataset page: https://huggingface.co/datasets/nguha/legalbench-staging.legalpincite
👩⚖️ LegalPincite: Multi-level Legal Information Retrieval Dataset
LegalPincite is a large-scale legal information retrieval (IR) test collection built from Court of Justice of the European Union (CJEU) judgments in EUR-Lex. It is designed for citation-oriented legal retrieval at multiple levels of granularity: case-to-case retrieval, paragraph-to-case retrieval, and paragraph-to-paragraph pinpoint citation (pincite) retrieval. The dataset is especially useful for… See the full description on the dataset page: https://huggingface.co/datasets/theresiavr/legalpincite.Tawan_legal_judgementLegalLensNER
Homepage: https://www.darrow.ai/
Repository: https://github.com/darrow-labs/LegalLens
Paper: https://arxiv.org/pdf/2402.04335.pdf
Point of Contact: Dor Bernsohn,Gil Semo
Overview
LegalLensNER is a dedicated dataset designed for Named Entity Recognition (NER) in the legal domain, with a specific emphasis on detecting legal violations in unstructured texts.
Data Fields
id: (int) A unique identifier for each record.
word: (str) The specific word or token in the… See the full description on the dataset page: https://huggingface.co/datasets/darrow-ai/LegalLensNER.tw-legal-benchmark-v1
Taiwan Legal Benchmark v1
A multiple-choice benchmark for evaluating large language models on Taiwan law in Traditional Chinese (繁體中文). It covers six legal domains with 209 questions drawn from Taiwan bar exam and certification-style questions.
Overview
Property
Value
Language
Traditional Chinese (zh-TW)
Questions
209
Format
4-choice multiple choice (A / B / C / D)
Domain
Taiwan law
License
Apache 2.0
Legal Domains Covered… See the full description on the dataset page: https://huggingface.co/datasets/lianghsun/tw-legal-benchmark-v1.singapore-legal-ai-benchmark
Singapore Legal AI Benchmark
Public research release of 102 Singapore legal research questions, model
responses from 6 systems, and overlapping grades on five dimensions.
Headline metrics are overlapping binary flags, not a ranking and not a
partition of 100%.
Interactive explorer
Open the explorer →
— comparison table, category heatmap, per-question comparison, and every answer
with its sources and grades.
(Space page)
Overall (n = 612)… See the full description on the dataset page: https://huggingface.co/datasets/JonathanSu/singapore-legal-ai-benchmark.Legal_Clause_Instructionslegal-contractslegal-qa-v1tw-legal-benchmark-v2
Taiwan Legal Benchmark v2
A multiple-choice benchmark for evaluating LLMs on Taiwan law in Traditional Chinese,
built from 15 years (2012–2026) of national examinations published by the
Ministry of Examination (考選部).
Supersedes tw-legal-benchmark-v1
(209 questions) with 17,002 deduplicated questions across 15 legal domains.
Overview
Property
Value
Questions
17,002 (deduplicated)
Years
2012–2026
Source papers
1,040 official exam papers
Format… See the full description on the dataset page: https://huggingface.co/datasets/lianghsun/tw-legal-benchmark-v2.legal-scenarios-SCOTUS-2024-decisions
Purpose and scope
This dataset evaluates an LLM's reasoning ability in a legal context. Each question presents a realistic scenario involving competing legal principals,
and asks the LLM to present a correct legal resolution with sufficient justification based on precedent. The dataset was created using slip opinions of
the US Supreme Court from the 2024 term, taken from the Supreme Court website.
Dataset Creation Method
The benchmark was created using RELAI’s data… See the full description on the dataset page: https://huggingface.co/datasets/relai-ai/legal-scenarios-SCOTUS-2024-decisions.indian-legalsynthetic-legal
⚖️ Synthetic Legal (Query, Response) Dataset
📚 140,000 synthetic (legal query, legal response) pairs across 13 legal domains, built to resemble the structure of real-world fact patterns and citation-backed answers.
⚠️ Disclaimer: All text is synthetically generated and IS NOT LEGALLY ACCURATE. Citations are real but assigned at random, and the verified_solution and verification_method columns are template labels, not evidence of review. This dataset is not legal advice.… See the full description on the dataset page: https://huggingface.co/datasets/Taylor658/synthetic-legal.swiss-legal-translation
Swiss Legal Translation Dataset
A large-scale parallel corpus of Swiss legal texts in German, French, and Italian, extracted from official government sources.
Dataset Overview
Metric
Value
Total articles
62,594
Languages
German, French, Italian
All 3 languages
28,000 articles (44.7%)
At least 2 languages
62,594 articles (100%)
Source laws
743+ unique laws
File size
~88 MB
Language Coverage by Source
Source
Articles
DE
FR
IT… See the full description on the dataset page: https://huggingface.co/datasets/liechticonsulting/swiss-legal-translation.portuguese-legal-sentences-v0
Work developed as part of Project IRIS.
Thesis: A Semantic Search System for Supremo Tribunal de Justiça
Portuguese Legal Sentences
Collection of Legal Sentences from the Portuguese Supreme Court of Justice
The goal of this dataset was to be used for MLM and TSDAE
Contributions
@rufimelo99
If you use this work, please cite:
@InProceedings{MeloSemantic,
author="Melo, Rui
and Santos, Pedro A.
and Dias, Jo{\~a}o",
editor="Moniz, Nuno
and Vale, Zita
and… See the full description on the dataset page: https://huggingface.co/datasets/stjiris/portuguese-legal-sentences-v0.LegalLensNLI
Homepage: https://www.darrow.ai/
Repository: https://github.com/darrow-labs/LegalLens
Paper: https://arxiv.org/pdf/2402.04335.pdf
Point of Contact: Dor Bernsohn,Gil Semo
Overview
The LegalLensNLI dataset is a unique collection of entries designed to show the connection between legal cases and the people affected by them. It's specially made for machine learning tools that aim to investigate more in the area of legal violations, specifically class action complaints. The main… See the full description on the dataset page: https://huggingface.co/datasets/darrow-ai/LegalLensNLI.turkish-legal-rag
Turkish Legal RAG Corpus — Türk Hukuku için Açık RAG Datasetı
Tek cümle: 25 önemli Türk kanununun (mevzuat.gov.tr kaynaklı, madde bazlı temiz chunk'lar) + 290 manuel doğrulanmış soru-cevap altın benchmark'ının olduğu açık kaynak Türkçe hukuk RAG datasetı.
🇹🇷 Türkçe Özet — Bu dataset, Türkçe hukuk uygulamaları için sıfırdan üretilmiş açık ve denetlenebilir bir RAG corpus'udur. mevzuat.gov.tr üzerinden alınan 25 ana kanunun madde madde temizlenmiş, chunk'lanmış sürümünü (6.350… See the full description on the dataset page: https://huggingface.co/datasets/mtntasci/turkish-legal-rag.legal_consolidationTask details
Legal consolidation is a critical yet time-consuming task, traditionally performed manually by legal professionals.
The objective is to automate the process of French legal consolidation, which is the application of modifications from a
modification section to an initial article to generate a modified article.
Dataset structure
A triplet of:
an initial article: the legislative article before consolidation,
a modification section: the text introducing the modification within… See the full description on the dataset page: https://huggingface.co/datasets/DoctrineAI/legal_consolidation.legalpincite
👩⚖️ LegalPincite: Multi-level Legal Information Retrieval Dataset
LegalPincite is a large-scale legal information retrieval (IR) test collection built from Court of Justice of the European Union (CJEU) judgments in EUR-Lex. It is designed for citation-oriented legal retrieval at multiple levels of granularity: case-to-case retrieval, paragraph-to-case retrieval, and paragraph-to-paragraph pinpoint citation (pincite) retrieval. The dataset is especially useful for… See the full description on the dataset page: https://huggingface.co/datasets/AccountVerify/legalpincite.legal_rag_hallucinations
Dataset Card for Hallucination Free? Assessing the Reliability of Leading AI Legal Research Tools
This data release contains the queries and raw model outputs we analyze in Magesh, Surani, Dahl, Suzgun, Manning and Ho, Hallucination Free? Assessing the Reliability of Leading AI Legal Research Tools, Journal of Empirical Legal Studies (2024, forthcoming).
Consistent with emerging understanding of AI
benchmarking and leaderboards, we reserve a random sample of 50% of the dataset to… See the full description on the dataset page: https://huggingface.co/datasets/reglab/legal_rag_hallucinations.legal-disclosure-coherence-breach-detection-v0.1What this dataset is
You receive
disclosure duty
material
timing
defence access
prejudice signals
You decide
Does disclosure behaviour match the legal duty
Answer
coherent
or
incoherent
Why this matters
Many unsafe convictions arise from disclosure failure.
This dataset measures the structural gap between duty and behaviour.
legalbench
Dataset Card for Dataset Name
Homepage: https://hazyresearch.stanford.edu/legalbench/
Repository: https://github.com/HazyResearch/legalbench/
Paper: https://arxiv.org/abs/2308.11462
Dataset Description
Dataset Summary
The LegalBench project is an ongoing open science effort to collaboratively curate tasks for evaluating legal reasoning in English large language models (LLMs). The benchmark currently consists of 162 tasks gathered from 40… See the full description on the dataset page: https://huggingface.co/datasets/Advocate512/legalbench.multilevel-legal-reasoning
Legal Reasoning Dataset with Multilevel Human and Model-Annotated Explanations
Prepared by Mst Rafia Islam, Umong Sain, Azmine Toushik Wasi
Prepared as a part of Reasoning Datasets Competition by Bespoke Labs, Hugging Face, and Together.ai.
🧭 Purpose and Scope
The Legal Reasoning Dataset aims to support the evaluation and training of legal reasoning systems, particularly in multilingual or jurisdiction-agnostic contexts. It focuses on international acts and treaties… See the full description on the dataset page: https://huggingface.co/datasets/ciol-research/multilevel-legal-reasoning.tw-legal-benchmark-v2
Taiwan Legal Benchmark v2
A multiple-choice benchmark for evaluating LLMs on Taiwan law in Traditional Chinese,
built from 15 years (2012–2026) of national examinations published by the
Ministry of Examination (考選部).
Supersedes tw-legal-benchmark-v1
(209 questions) with 17,002 deduplicated questions across 15 legal domains.
Overview
Property
Value
Questions
17,002 (deduplicated)
Years
2012–2026
Source papers
1,040 official exam papers
Format… See the full description on the dataset page: https://huggingface.co/datasets/Jel1f1sh/tw-legal-benchmark-v2.sen_legal_graphrag_dataLegalSeg_CSVlegal_reason
Enhanced Legal Reasoning Dataset
Dataset Description
Dataset Summary
The Enhanced Legal Reasoning Dataset is a synthetic dataset designed to facilitate the fine-tuning of Large Language Models (LLMs) for tasks related to legal reasoning and argumentation. It encompasses a diverse range of legal scenarios across multiple domains, capturing the nuanced techniques employed by legal professionals in constructing their arguments.
Dataset Structure
The… See the full description on the dataset page: https://huggingface.co/datasets/chemouda/legal_reason.legal_summary