SciEvalKit: An Open-source Evaluation Toolkit for Scientific General Intelligence
IntermediateYiheng Wang, Yixin Chen et al.Dec 26arXiv
SciEvalKit is a new open-source toolkit that tests AI on real scientific skills, not just trivia or simple Q&A.
#scientific intelligence evaluation#multimodal scientific reasoning#symbolic reasoning in AI