Published in arXiv, 2025
We present EmbodiedBrain, a large-scale embodied planning model trained with multimodal post-training and reinforcement learning.
Recommended citation: Ding Zou et al. EmbodiedBrain: Expanding Performance Boundaries of Task Planning for Embodied Intelligence. arXiv:2510.20578.
Published in AAAI 2026, 2026
This paper revisits multimodal data sampling strategies for reinforcement learning based post-training from a difficulty-distinguish perspective.
Recommended citation: J Qi, Ding Zou, Wentao Yan, Rui Ma, Jiawei Li, Zhibin Zheng, Zhen Yang, and Rui Zhao. Revisiting the Data Sampling in Multimodal Post-training from a Difficulty-Distinguish View. AAAI 2026.
Published in EMNLP 2025 Findings, 2025
We propose Curriculum Reinforcement Learning for improving reasoning and out-of-distribution generalization of small-scale Vision-Language Models.
Recommended citation: Ding Zou et al. Boosting the Generalization and Reasoning of Vision Language Models with Curriculum Reinforcement Learning. EMNLP 2025 Findings.
Published in WWW 2024, 2024
This paper addresses the challenges of multi-intent modeling and knowledge noise in knowledge-aware recommendation systems. We propose a knowledge-enhanced multi-intent transformer that integrates global heterogeneous information to model user intents while selectively filtering intent-irrelevant knowledge triples. Extensive experiments on multiple benchmark datasets demonstrate consistent improvements over strong baselines.
Recommended citation: Ding Zou, Wei Wei, Feida Zhu, Chuanyu Xu, Tao Zhang, and Chengfu Huo. Knowledge Enhanced Multi-intent Transformer Network for Recommendation. Companion Proceedings of the ACM Web Conference 2024, 1-9.
Published in SIGIR 2022, 2022
We study the sparse supervision problem in knowledge-aware recommender systems and introduce a self-supervised contrastive learning framework. Our approach models multiple graph views at different semantic levels and performs cross-view contrastive learning to enhance representation robustness. Experimental results show consistent gains across multiple public benchmarks.
Recommended citation: Ding Zou, Wei Wei, Xian-Ling Mao, Ziyang Wang, Minghui Qiu, Feida Zhu, and Xin Cao. Multi-level Cross-view Contrastive Learning for Knowledge-aware Recommender System. SIGIR 2022, 1358-1368.