---
title: README
emoji: 📘
colorFrom: blue
colorTo: gray
sdk: static
pinned: false
---
 
<p>
  Welcome to the official Hugging Face organization for Oxford University Press' open benchmarks! 🤗
</p>
<p>
  <strong>Open benchmarks:</strong><br>
</p>
<ul>
  <li><a href="https://huggingface.co/datasets/OUP/l2-bench" class="underline">L2-Bench</a>: our public benchmark dataset for measuring LLM capabilities in Second Language (L2) Education.</li>
</ul>
<p>
  <strong>Resources:</strong>
</p>
<ul>
  <li><a href="https://huggingface.co/spaces/OUP/l2-bench-pipeline/blob/main/README.md" class="underline">L2-Bench Pipeline</a>: our reference implementation for L2-Bench, involving response generation and LLM-as-judge scoring pipeline with `inspect-ai`.</li>
  <li><a href="https://arxiv.org/abs/2607.08842" class="underline">L2-Bench Paper</a>: our technical paper, "L2-Bench: An Evaluation Benchmark for Measuring LLM Capabilities in Second Language Education".</li>
  <li><a href="https://benchmarks.elt.edu.oup.com" class="underline">L2-Bench Site</a>: our L2-Bench site and blog (for any inquiries or feedback, please contact us via the "Register Interest" form).</li>
</ul>