Autoregressive robot video world models based on discrete visual tokens remain unreliable over long rollout horizons. Retaining the entire generated history does not necessarily improve prediction quality. We investigate a possible memory-role conflict in Full Context decoding: persistent scene information, recent visu...
Hao Wu, Tian-Yu Shao, Sheng-Ju Qian et al.· 0 citations
The model achieves the best performance among similarly sized models on 19 of the 38 benchmarks and substantially outperforms strong competitors, including Qwen3.6-A3B and Cosmos 3.5 MoT-2B, and demonstrates strong performance on embodied agentic tasks requiring multi-turn interaction and long-horizon reasoning.
This work shows that a single BDDL specification, automatically constructed from open-world video evidence or curated tasks, can serve as a shared interface for data construction, plan verification, and reward design and introduces GroupAdapt, a difficulty-aware length schedule that uses the in-batch group pass rate as...
Xiangli Shi, Xiaomeng Zhu, Ye Tian 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.
