This work explores model performance on multi-step world state prediction, a high-level task that requires models to forecast successive future world states from multimodal observations and diverse references, and highlights the need for multimodal generative models that are capable of dynamic planning and reasoning wh...
Wen-Hao Yan, Hao-Dong Li, Xiao-Tong Li et al.· 1 citation
Transforming multimodal sources into condensed and structured media outputs can be fundamentally conceptualized as a long-horizon agentic process centered on a model-harness system. While an ideal harness system should align with human design priors and accumulate reusable experience through empirical exploration to dr...
Yaxin Luo, Hao-Bin Jiang, Jialv Zou et al.· 6 citations
FigMirror is proposed, an agentic framework that unlocks coordinate grounding and coding capabilities present in modern computer-use models through Grounded Measurement, which locates visual elements by coordinates and measures their properties through executable code.
Xiaohan Zhao, Jiacheng Liu, Yaxin Luo et al.· 0 citations
VideoCoCo, an agentic dual-engine framework in which executable Blender code serves as a process-level chain of thought, demonstrates that executable code provides an effective, controllable, and inspectable intermediate representation for physically consistent video generation.
Haodong Li, Tianfei Ren, Xiaoxiao Ma et al.· arXiv.org· 8 citations
DARTree is introduced, a training-free speculative decoding method that extends a pretrained AR correction head from chains to trees, and achieves the highest average acceptance length and speedup in all four model--temperature configurations.