Ye, Yiran, Thai Le, and Dongwon Lee. 2025. “NoisyHate: Mining Online Human-Written Perturbations for Realistic Robustness Benchmarking of Content Moderation Models”. Proceedings of the International AAAI Conference on Web and Social Media 19 (1):2603-12. https://doi.org/10.1609/icwsm.v19i1.35961.