General-purpose multimodal foundation models are increasingly capable of directly controlling robots, yet improving their manipulation capabilities through autonomous interaction remains challenging. A key challenge to such self-improvement is learning from failures in complex manipulation tasks involving cross-stage d...
This work introduces VideoResearchAgent, a scalable training framework for open-web video research, and introduces Retrieval-Domain-Randomized GRPO (RDR-GRPO), which diversifies candidate rankings, distractors, metadata, and result structure during training to reduce overfitting to simulated retrieval.
Yu-Hang Zhou, Fei-Yu Li, Yu-Xin Wu et al.· 0 citations
Vision-Language-Action (VLA) models have shown strong capabilities in robotic manipulation, yet existing benchmarks typically assume that task-relevant information is explicitly specified in the instruction. In practice, however, humans frequently refer to objects, quantities, and relations implicitly, requiring robots...
Aernaer Akelijiang, Jian-Nan Li, Zhi-Neng Chen et al.· 0 citations
Robotic foundation models achieve impressive performance on standard manipulation benchmarks, yet these evaluations typically assume clean, timely, and consistent visual observations throughout execution. We introduce LIBERO-VPro, a benchmark for systematically evaluating the closed-loop visual robustness of robotic fo...
Hui-Qiong Li, Zhi-Ting Mei, Anirudha Majumdar et al.· 1 citation
Overall, RAPolicy substantially outperforms the baselines in aggregate task success while requiring fewer human interventions, demonstrating stable policy improvement and high online training efficiency.
Jia-Rui Yang, Jia-Jin Zhang, Bin Zhu et al.· 0 citations
Experiments demonstrate that MoWAM achieves strong in-distribution performance, improved out-of-distribution robustness, and higher average real-world success than representative WAM baselines, demonstrating that explicit future motion provides an effective and efficient basis for inference-time scaling.
VICAL is introduced, a consistency-driven framework that improves long-tailed recognition not by enforcing expert diversity, but by reducing prediction variance, and suggests that multi-expert models benefit more from variance reduction than diversity maximization.
Jianggang Zhu, Zheng Wang, Bin Zhu et al.· 0 citations
Across three flow-based VLA models on multiple simulated manipulation benchmarks and two real-world tasks, StructRL improves exploration efficiency and OOD performance over prior in-chain baselines, demonstrating the effectiveness of structured action-space exploration for adapting flow-based VLA with RL.
Jia-Rui Yang, Bin Zhu, Jing-Jing Chen et al.· 1 citation
Saliency-guided Purification and Adaptive Redistribution (SPAR), a training-free, plug-and-play intervention that mitigates this generalized textual bias exerted over visual features that extends beyond isolated sink tokens.
Peng-Kun Jiao, Bin Zhu, Jingjing Chen et al.· arXiv.org· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
