ConsiSpace is proposed, a geometry-consistency-aware framework for geometry-sensitive video spatial reasoning that turns spatial consistency into both an evidence organization principle and an explicit post-SFT learning signal, and utilizes unified consistency self-supervised reinforcement learning (UC-SSRL) after supervised fine-tuning to improve cross-view stability.
DeCoRAG is a multimodal Graph RAG pipeline that shifts knowledge processing from coupled visual-semantic reasoning to cognitive Decoupling, and subsequently drives the reasoning space from dense, noisy backgrounds to purified, intent-driven semantic clusters.
Shuoshuo Wang, Kai Zhang, Wenyuan Huang et al.· arXiv.org· 0 citations
This work proposes Knowledge Access Planning (KAP), a paradigm-shifting execution abstraction that elevates structured knowledge priors from passive prompt-construction hints into first-class physical execution artifacts, and derives a phase-boundary model for the positive-speedup regime of plan-guided execution.