ManimAgent: Self-Evolving Multimodal Agents for Visual Education
Published in arXiv, 2026
ManimAgent is a self-evolving multimodal agent for generating mathematical teaching animations with Manim. Its dual-channel Episodic Memory Bank carries successful patterns and validated failure lessons across tasks without weight updates or human-seeded examples.
| Project page | Code | Google Scholar |
Recommended citation: Wenjia Jiang, Zongyuan Cai, Yuanhang Shao, Chenru Wang, Boyan Han, Zhixue Song, Keyu Chen, Shengwei An, Xu Yang, Zhou Yang. (2026). "ManimAgent: Self-Evolving Multimodal Agents for Visual Education." arXiv preprint arXiv:2606.30296.
Download Paper