Results indicate that minimally sufficient world representations provide an effective foundation for robust VLN, and propose BrainNav, a navigation framework grounded in the Principle of Minimal Sufficiency.
LightNav-0 is presented, a compact generalist embodied navigation model that elicits the spatial intelligence of a pretrained VLM and aligns it with navigation, without task-specific prediction heads, and establishes compact VLMs as a unified and transferable backbone for generalist embodied navigation.
The real-robot benchmark demonstrates that StellaVLA can use both human/robot demos and human-to-robot (XR) demos as in-context structured demonstration to help VLA model adapt to OOD tasks.
This work proposes GroundingVLN, which uses visual grounding as a shared interface between reasoning and action, and introduces Grounded and Execution-Aware Reinforcement Learning (GEAR), which aligns grounded reasoning and spatial decisions with downstream execution.
This work proposes TAMP-Nav, a unified framework for efficient embodied navigation that dynamically triggers Chain-of-Thought and retains high-fidelity memory only at critical nodes, compressing redundant trajectories into lightweight Space-Time Indicators, thereby preserving critical historical information and enhanci...
AnchorGUI is proposed, a unified framework driven by the Cognitive State Anchor, which acts as a per-step primitive that actively compares expected and observed transitions, converting passive multimodal trajectories into explicit prediction-error signals that orchestrate a dual-scale learning mechanism via an asymmetr...
We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.
