AGIBench: A Multi-granularity, Multimodal, Human-referenced, Auto-scoring Benchmark for Large Language Models

no code implementations • 5 Sep 2023 • Fei Tang, Wanling Gao, Luzhou Peng, Jianfeng Zhan

Instead of a collection of blended questions, AGIBench focuses on three typical ability branches and adopts a four-tuple <ability branch, knowledge, difficulty, modal> to label the attributes of each question.

Benchmarking Zero-Shot Learning