A unified evaluation toolkit and leaderboard for rigorously assessing the scientific intelligence of large language and vision–language models across the full research workflow. Maintained by InternScience on GitHub, where it has earned 85 stars from the community.
