Shypd.aiThe largest open index of AI toolsGAOKAO-Bench is an evaluation framework that utilizes GAOKAO questions as a dataset to evaluate large language models. It helps assess language understanding and logical reasoning abilities of AI models.
Trending
About
GAOKAO-Bench is an evaluation framework designed to assess the language understanding and logical reasoning capabilities of large language models (LLMs). It leverages questions from the Chinese GAOKAO exam, a highly standardized and comprehensive test, as its dataset. The framework includes 1781 objective questions and 1030 subjective questions from GAOKAO exams between 2010 and 2022. It provides methods for evaluating models in a zero-shot setting, including rule-based answer extraction for objective questions and human or LLM-as-a-Judge scoring for subjective questions. The platform also offers tools to calculate overall GAOKAO scores for various LLMs, providing a standardized benchmark for comparison.
Capabilities
Pricing & Plans
Open Source
Free
FAQs
Trending
.png&w=3840&q=75){kind=logo}
