Please download the full version of the dataset from Zenodo, here.

Contract Understanding Atticus Dataset (CUAD) v1 is a corpus of more than 13,000 labels in 510 commercial legal contracts that have been manually labeled by The Atticus Project to identify 41 categories of important clauses that lawyers look for when reviewing contracts.

We tested CUAD v1 against ten pretrained AI models and published the results on arXiv here.

Code for replicating the results, together with the model trained on CUAD, is published on Github here.

Contract Understanding Atticus Dataset (CUAD) is a dataset for legal contract review. CUAD was created with dozens of legal experts from The Atticus Project and consists of over 13,000 annotations. The task is to highlight salient portions of a contract that are important for a human to review.

Dataset Card for filtered_cuad

  Dataset Summary

Contract Understanding Atticus Dataset (CUAD) v1 is a corpus of more than 13,000 labels in 510 commercial legal contracts that have been manually labeled to identify 41 categories of important clauses that lawyers look for when reviewing contracts in connection with corporate transactions. This dataset is a filtered version of CUAD. It excludes legal contracts with an Agreement date prior to 2002 and contracts which are not… See the full description on the dataset page: https://huggingface.co/datasets/alex-apostolo/filtered-cuad.

Dataset Card for CUAD

This is a modified version of original CUAD which trims the question to its label form.

  Dataset Summary

Contract Understanding Atticus Dataset (CUAD) v1 is a corpus of more than 13,000 labels in 510 commercial legal contracts that have been manually labeled to identify 41 categories of important clauses that lawyers look for when reviewing contracts in connection with corporate transactions. CUAD is curated and maintained by The Atticus Project, Inc.… See the full description on the dataset page: https://huggingface.co/datasets/chenghao/cuad_qa.

CUAD-DeepSeek: Enhanced Legal Contract Understanding Dataset

CUAD-DeepSeek is an enhanced version of the Contract Understanding Atticus Dataset (CUAD), enriched with expert rationales and reasoning traces provided by the DeepSeek language model. This dataset aims to improve legal contract analysis by providing not just classifications but detailed explanations for why specific clauses belong to particular legal categories.

  Purpose and Scope

Legal contract review is… See the full description on the dataset page: https://huggingface.co/datasets/zenml/cuad-deepseek.

Dataset Card for Contract Understanding Atticus Dataset (CUAD) PDF

This dataset contains the PDFs and the full text of 509 commercial legal contracts from the original CUAD dataset. One of the original 510 contracts was removed due to being a scanned copy. The extracted text was cleaned using clean-text. The PDFs were encoded in base64 and added as the pdf_bytes_base64 feature. You can easily and quickly load it: dataset = load_dataset("dvgodoy/CUAD_v1_Contract_Understanding_PDF")… See the full description on the dataset page: https://huggingface.co/datasets/dvgodoy/CUAD_v1_Contract_Understanding_PDF.

Merger Agreement Understanding Dataset (MAUD) v1 is a corpus of 47,000+ labels in 152 merger agreements that have been manually labeled under the supervision of experienced lawyers to identify 92 questions in each agreement used by the 2021 American Bar Association (ABA) Public Target Deal Points Study.

MAUD is curated and maintained by The Atticus Project, Inc. to support NLP research and development in legal contract review.

ReadMe and Datasheet are published here. Code for replicating the results, together with the model trained on CUAD, is published on Github here.

CUADGoverningLawLegalBenchClassification An MTEB dataset Massive Text Embedding Benchmark

This task was constructed from the CUAD dataset. It consists of determining if the clause specifies which state/country’s law governs the contract.

Task category t2c

Domains Legal, Written

Reference https://huggingface.co/datasets/nguha/legalbench

  How to evaluate on this task

You can evaluate an embedding model on this dataset using the following code: import mteb… See the full description on the dataset page: https://huggingface.co/datasets/mteb/CUADGoverningLawLegalBenchClassification.

CUADInsuranceLegalBenchClassification An MTEB dataset Massive Text Embedding Benchmark

This task was constructed from the CUAD dataset. It consists of determining if clause creates a requirement for insurance that must be maintained by one party for the benefit of the counterparty.

  Task category

t2c

Domains Legal, Written

Reference https://huggingface.co/datasets/nguha/legalbench

  How to evaluate on this task

You can evaluate an embedding model on this… See the full description on the dataset page: https://huggingface.co/datasets/mteb/CUADInsuranceLegalBenchClassification.

Merger Agreement Understanding Dataset (MAUD) v1 is a corpus of 47,000+ labels in 152 merger agreements that have been manually labeled under the supervision of experienced lawyers to identify 92 questions in each agreement used by the 2021 American Bar Association (ABA) Public Target Deal Points Study.

MAUD is curated and maintained by The Atticus Project, Inc. to support NLP research and development in legal contract review.

ReadMe and Datasheet are published here. Code for replicating the results, together with the model trained on CUAD, is published on Github here.

MAUD is an expert-annotated merger agreement reading comprehension dataset based on the American Bar Association's 2021 Public Target Deal Points study, where lawyers and law students answered 92 questions about 152 merger agreements.

With over 39,000 examples and 47,000 total annotations, it is the largest expert-annotated legal reading comprehension dataset in the English language, as well as the first expert-annotated merger agreement dataset.

CUADAntiAssignmentLegalBenchClassification An MTEB dataset Massive Text Embedding Benchmark

This task was constructed from the CUAD dataset. It consists of determining if the clause requires consent or notice of a party if the contract is assigned to a third party.

Task categoryt2c

Domains Legal, Written

Reference https://huggingface.co/datasets/nguha/legalbench

  How to evaluate on this task

You can evaluate an embedding model on this dataset using the… See the full description on the dataset page: https://huggingface.co/datasets/mteb/CUADAntiAssignmentLegalBenchClassification.

CUADWarrantyDurationLegalBenchClassification An MTEB dataset Massive Text Embedding Benchmark

This task was constructed from the CUAD dataset. It consists of determining if the clause specifies a duration of any warranty against defects or errors in technology, products, or services provided under the contract.

Task category t2c

Domains Legal, Written

Reference https://huggingface.co/datasets/nguha/legalbench

  How to evaluate on this task

You can evaluate an… See the full description on the dataset page: https://huggingface.co/datasets/mteb/CUADWarrantyDurationLegalBenchClassification.

CUADLicenseGrantLegalBenchClassification An MTEB dataset Massive Text Embedding Benchmark

This task was constructed from the CUAD dataset. It consists of determining if the clause contains a license granted by one party to its counterparty.

Task category t2c

Domains Legal, Written

Reference https://huggingface.co/datasets/nguha/legalbench

  How to evaluate on this task

You can evaluate an embedding model on this dataset using the following code: import… See the full description on the dataset page: https://huggingface.co/datasets/mteb/CUADLicenseGrantLegalBenchClassification.

CUADThirdPartyBeneficiaryLegalBenchClassification An MTEB dataset Massive Text Embedding Benchmark

This task was constructed from the CUAD dataset. It consists of determining if the clause specifies that that there a non-contracting party who is a beneficiary to some or all of the clauses in the contract and therefore can enforce its rights against a contracting party.

Task category t2c

Domains Legal, Written

Reference https://huggingface.co/datasets/nguha/legalbench… See the full description on the dataset page: https://huggingface.co/datasets/mteb/CUADThirdPartyBeneficiaryLegalBenchClassification.

CUADMostFavoredNationLegalBenchClassification An MTEB dataset Massive Text Embedding Benchmark

This task was constructed from the CUAD dataset. It consists of determining if a third party gets better terms on the licensing or sale of technology/goods/services described in the contract, the buyer of such technology/goods/services under the contract shall be entitled to those better terms.

Task category t2c

Domains Legal, Written

Reference… See the full description on the dataset page: https://huggingface.co/datasets/mteb/CUADMostFavoredNationLegalBenchClassification.

CUADRevenueProfitSharingLegalBenchClassification An MTEB dataset Massive Text Embedding Benchmark

This task was constructed from the CUAD dataset. It consists of determining if the clause require a party to share revenue or profit with the counterparty for any technology, goods, or services.

Task category t2c

Domains Legal, Written

Reference https://huggingface.co/datasets/nguha/legalbench

  How to evaluate on this task

You can evaluate an embedding… See the full description on the dataset page: https://huggingface.co/datasets/mteb/CUADRevenueProfitSharingLegalBenchClassification.

CUADExpirationDateLegalBenchClassification An MTEB dataset Massive Text Embedding Benchmark

This task was constructed from the CUAD dataset. It consists of determining if the clause specifies the date upon which the initial term expires.

Task category t2c

Domains Legal, Written

Reference https://huggingface.co/datasets/nguha/legalbench

  How to evaluate on this task

You can evaluate an embedding model on this dataset using the following code: import mteb… See the full description on the dataset page: https://huggingface.co/datasets/mteb/CUADExpirationDateLegalBenchClassification.

CUADPriceRestrictionsLegalBenchClassification An MTEB dataset Massive Text Embedding Benchmark

This task was constructed from the CUAD dataset. It consists of determining if the clause places a restriction on the ability of a party to raise or reduce prices of technology, goods, or services provided.

Task category t2c

Domains Legal, Written

Reference https://huggingface.co/datasets/nguha/legalbench

  How to evaluate on this task

You can evaluate an… See the full description on the dataset page: https://huggingface.co/datasets/mteb/CUADPriceRestrictionsLegalBenchClassification.

CUADVolumeRestrictionLegalBenchClassification An MTEB dataset Massive Text Embedding Benchmark

This task was constructed from the CUAD dataset. It consists of determining if the clause specifies a fee increase or consent requirement, etc. if one party's use of the product/services exceeds certain threshold.

Task category t2c

Domains Legal, Written

Reference https://huggingface.co/datasets/nguha/legalbench

  How to evaluate on this task

You can evaluate an… See the full description on the dataset page: https://huggingface.co/datasets/mteb/CUADVolumeRestrictionLegalBenchClassification.