Spaces:
Sleeping
Sleeping
File size: 759 Bytes
ee81350 288fdbf ee81350 288fdbf 4126a18 288fdbf ee81350 288fdbf |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 |
---
title: ATLAS Benchmark
emoji: 🧪
colorFrom: green
colorTo: indigo
sdk: gradio
app_file: app.py
pinned: true
license: apache-2.0
short_description: ATLAS for Frontier Scientific Benchmark
sdk_version: 5.43.1
hf_oauth: true
tags:
- leaderboard
- science
- benchmark
- evaluation
---
# ATLAS: A High-Difficulty, Multidisciplinary Benchmark for Frontier Scientific Reasoning
ATLAS is a high-difficulty, multidisciplinary benchmark for frontier scientific reasoning. It is designed to evaluate the capabilities of large language models (LLMs) in scientific reasoning across seven core scientific fields covering the key domains of AI for Science (AI4S):
- Mathematics
- Physics
- Chemistry
- Biology
- Computer Science
- Earth Science
- Materials Science
|