Datasets › LegalBench

LegalBench

Introduced by Neel Guha et al. in LegalBench: A Collaboratively Built Benchmark for Measuring Legal Reasoning in Large Language Models20 Aug 2023 archive 2025-07-28

LegalBench is a fascinating project that revolves around legal reasoning and evaluation. Let me break it down for you:

  1. LegalBench Website:

    • The LegalBench website serves as your ultimate legal helper. It's a platform where you can hire expert lawyers for various legal tasks, all online and from anywhere.
    • Here's how it works:
      1. Post a Job: If you have a legal task that needs completion, simply post the job on LegalBench.
      2. Receive Bids: Within minutes, you'll start receiving competitive bids from freelancers who can handle your legal project.
      3. Connect with Lawyers: Choose the right lawyer for your needs and get your project done.
    • LegalBench covers a range of legal categories, including patents and legal advice¹.
  2. LegalBench Benchmark:

    • 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 contributors.
    • Each task in the benchmark has an associated dataset, consisting of input-output pairs.
    • It's a valuable resource for assessing how well language models understand and reason about legal concepts and scenarios.
    • If you're curious, you can explore the LegalBench GitHub repository for more details²³.

Source: Conversation with Bing, 3/16/2024 (1) LegalBench | Your ultimate legal helper. https://legalbench.com/. (2) LegalBench - GitHub. https://github.com/HazyResearch/legalbench. (3) Home | LegalBench. https://hazyresearch.stanford.edu/legalbench/. (4) LegalBench Dataset | Papers With Code. https://paperswithcode.com/dataset/legalbench.

Benchmarks archive 2025-07-28

No leaderboard in the archive resolves to this dataset.

Papers archive 2025-07-28

No paper in the archive has a leaderboard row on this dataset; the archive counts 40 papers for it but never published that list.

Dataset loaders archive 2025-07-28

No loader listed in the archive.

Tasks archive 2025-07-28

No task tagged in the archive.

License archive 2025-07-28

No licence recorded in the archive. Absence here is not a statement about the dataset's terms.

Modalities archive 2025-07-28

No modality tagged.

Languages archive 2025-07-28

Variants archive 2025-07-28

  • LegalBench

1 variant name, as the archive lists them.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections