ARC
Emerging2papers using it
2025first seen
The ARC dataset/benchmark contains a variety of tasks designed to evaluate the reasoning abilities of models, particularly in the context of multiple-choice questions.
The ARC dataset/benchmark contains a variety of tasks designed to evaluate the reasoning abilities of models, particularly in the context of multiple-choice questions.