---
license: apache-2.0
---

This is benchmark for paper "ActiView: Evaluating Active Perception Ability for Multimodal Large Language Models" ([arxiv](https://arxiv.org/abs/2410.04659))

Please refer to this [github](https://github.com/THUNLP-MT/ActiView) repo for detail.