xw27/scibench download history

xw27/scibench is a dataset on the Hugging Face Hub. In the last 30 days it was downloaded 6,872 times (1,736 in the last 7 days), and 34,693 times in total. It ranks #4,342 among datasets by monthly downloads.

SciBench SciBench is a novel benchmark for college-level scientific problems sourced from instructional textbooks. The benchmark is designed to evaluate the complex reasoning capabilities, strong domain knowledge, and advanced calculation skills of LLMs. Please refer to our paper or webs

Models trained on scibench

1 models list it as training data.

Spaces using scibench

Open xw27/scibench on Hugging Face