Extended Data Table 1 Overview of sources of the curated questions

From: A framework for evaluating the chemical knowledge and reasoning abilities of large language models against the expertise of chemists

Source

Count

Semi-automatically generated

1749

URL

375

Textbook

206

Exam

149

IChO

149

No source

139

Lectures

21

  1. The table provides an overview of the types of sources the questions have been curated from. Detailed sources are available in the source data on GitHub. Questions without a source have been curated completely from scratch. Questions based on lecture notes or URLs have been curated based on content presented in those resources. All questions have been rephrased, annotated, and reviewed before being added to the corpus.
Back to article page