TY  - RPRT
TI  - Aetheria: A multimodal interpretable content safety framework based on multi-agent debate and collaboration
AU  - Yuxiang He
AU  - Jian Zhao
AU  - Yuchen Yuan
AU  - Tianle Zhang
AU  - Wei Cai
AU  - Haojie Cheng
AU  - Ziyan Shi
AU  - Ming Zhu
AU  - Haichuan Tang
AU  - Chi Zhang
AU  - Xuelong Li
PY  - 2025
UR  - https://arxiv.org/abs/2512.02530
ID  - 2512.02530
ER  -