Saved toolsSubmit a tool
AI TOOL

A comprehensive large model assessment benchmark in Chinese

Claim toolCompare toolsReport a correction—
SHARE / FOXAI

Share tool · CMMLU

What is CMMLU?

CMMLU: Measuring massive multitask language understanding in Chinese - haonan-li/CMMLU

CMMLU IS A COMPREHENSIVE ASSESSMENT BENCHMARK IN CHINESE, DEDICATED TO ASSESSING THE KNOWLEDGE AND REASONING OF LANGUAGE MODELS IN THE CHINESE LANGUAGE CONTEXT, COVERING 67 TOPICS RANGING FROM BASIC DISCIPLINES TO ADVANCED PROFESSIONAL LEVELS. INCLUDING NATURAL SCIENCES REQUIRING COMPUTATION AND REASONING, HUMANITIES AND SOCIETIES REQUIRING KNOWLEDGE

简体中文 | English

Key features

  • Highlight

  • Dataset

  • Preprocessing code

  • Assessment tool

Use cases and audience

To be edited for verification

  • Language model performance assessment
  • Smart counselling in the field of education
  • Smart client service optimization
  • Cultural knowledge dissemination
Input

To be edited for verification

Output

To be edited for verification

You may also like

View all

The Wenxin AgentBuilder platform is Baidu’s agent-building platform powered by the Wenxin large model. Developers can choose their development approach based on industry and application scenarios, define agent capabilities through prompt orchestration, and reach users via the distribution channels provided by the platform.

View details