Mind over Space: Can Multimodal Large Language Models Mentally Navigate?

• URL: http://arxiv.org/abs/2603.21577v1
• Date: Mon, 23 Mar 2026 04:59:25 GMT
• Title: Mind over Space: Can Multimodal Large Language Models Mentally Navigate?
• Abstract summary: Mental navigation is the strategic construction of spatial representations from experience.<n>Video2 Frontier is a pioneering benchmark for evaluating the mental navigation capabilities of MLLMs.<n>Experiments demonstrate that NavMind achieves superior mental navigation capabilities.
• Score: 45.1484159975388
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Despite the widespread adoption of MLLMs in embodied agents, their capabilities remain largely confined to reactive planning from immediate observations, consistently failing in spatial reasoning across extensive spatiotemporal scales. Cognitive science reveals that Biological Intelligence (BI) thrives on "mental navigation": the strategic construction of spatial representations from experience and the subsequent mental simulation of paths prior to action. To bridge the gap between AI and BI, we introduce Video2Mental, a pioneering benchmark for evaluating the mental navigation capabilities of MLLMs. The task requires constructing hierarchical cognitive maps from long egocentric videos and generating landmark-based path plans step by step, with planning accuracy verified through simulator-based physical interaction. Our benchmarking results reveal that mental navigation capability does not naturally emerge from standard pre-training. Frontier MLLMs struggle profoundly with zero-shot structured spatial representation, and their planning accuracy decays precipitously over extended horizons. To overcome this, we propose \textbf{NavMind}, a reasoning model that internalizes mental navigation using explicit, fine-grained cognitive maps as learnable intermediate representations. Through a difficulty-stratified progressive supervised fine-tuning paradigm, NavMind effectively bridges the gap between raw perception and structured planning. Experiments demonstrate that NavMind achieves superior mental navigation capabilities, significantly outperforming frontier commercial and spatial MLLMs.

Related papers

• SpaceVLN: A Zero-Shot Vision-and-Language Navigation Agent with Online Spatial Cognitive Memory and Reasoning [59.64305326980364]
  SpaceVLN is a navigation agent built around Spatial Cognitive Memory and Task-Guided Spatial Reasoning.<n>Built on this memory, Spatial-CoT integrates task-progress reasoning with spatial perception, analysis, and prediction.<n>Across R2R-CE, RxR-CE, GN-Bench, and HM3D-OVON, SpaceVLN achieves state-of-the-art zero-shot performance.
  arXiv  Detail & Related papers  (2026-06-08T03:42:08Z)
• AgentVLN: Towards Agentic Vision-and-Language Navigation [78.739525400071]
  Vision-and-Language Navigation (VLN) requires an embodied agent to ground complex natural-language instructions into long-horizon navigation in unseen environments.<n>In this paper, we propose AgentVLN, a novel and efficient embodied navigation framework that can be deployed on edge computing platforms.
  arXiv  Detail & Related papers  (2026-03-18T12:43:47Z)
• VLingNav: Embodied Navigation with Adaptive Reasoning and Visual-Assisted Linguistic Memory [43.2995099083993]
  VLA models have shown promising potential in embodied navigation by unifying perception and planning.<n>Most existing VLA models rely on reactive mappings directly from observations to actions.<n>We propose VLingNav, a VLA model for embodied navigation grounded in linguistic-driven cognition.
  arXiv  Detail & Related papers  (2026-01-13T15:43:43Z)
• SpatialDreamer: Incentivizing Spatial Reasoning via Active Mental Imagery [64.67498968405327]
  SpatialDreamer is a reinforcement learning framework that enables spatial reasoning through a closedloop process of active exploration.<n>GeoPO introduces tree-structured sampling and step-level reward estimation with consistency geometric constraints.
  arXiv  Detail & Related papers  (2025-12-08T17:20:50Z)
• Think, Remember, Navigate: Zero-Shot Object-Goal Navigation with VLM-Powered Reasoning [8.316876835702525]
  Vision-Language Models (VLMs) are set to transform robotic navigation.<n>Our framework outsources high-level planning to a VLM, which leverages its contextual understanding to guide a frontier-based exploration agent.<n>When tested on challenging benchmarks like HM3D, Gibson, and MP3D, this method produces exceptionally direct and logical trajectories.
  arXiv  Detail & Related papers  (2025-11-12T03:38:50Z)
• Mind Meets Space: Rethinking Agentic Spatial Intelligence from a Neuroscience-inspired Perspective [53.556348738917166]
  Recent advances in agentic AI have led to systems capable of autonomous task execution and language-based reasoning.<n>Human spatial intelligence, rooted in integrated multisensory perception, spatial memory, and cognitive maps, enables flexible, context-aware decision-making in unstructured environments.
  arXiv  Detail & Related papers  (2025-09-11T05:23:22Z)
• From reactive to cognitive: brain-inspired spatial intelligence for embodied agents [50.99942960312313]
  Brain-inspired Spatial Cognition for Navigation (BSC-Nav) is a unified framework for constructing and leveraging structured spatial memory in embodied agents.<n> BSC-Nav builds allocentric cognitive maps from egocentric trajectories and contextual cues, and dynamically retrieves spatial knowledge aligned with semantic goals.
  arXiv  Detail & Related papers  (2025-08-24T03:20:48Z)
• DORAEMON: Decentralized Ontology-aware Reliable Agent with Enhanced Memory Oriented Navigation [55.888688171010365]
  DORAEMON is a cognitive-inspired framework consisting of Ventral and Dorsal Streams that mimics human navigation capabilities.<n>We evaluate DORAEMON on the HM3D, MP3D and GOAT datasets, where it achieves state-of-the-art performance on both success rate (SR) and success weighted by path length (SPL) metrics.
  arXiv  Detail & Related papers  (2025-05-28T04:46:13Z)
• Endowing Embodied Agents with Spatial Reasoning Capabilities for Vision-and-Language Navigation [0.0]
  We propose BrainNav, a bio-inspired spatial cognitive navigation framework inspired by biological spatial cognition theories and cognitive map theory.<n>BrainNav integrates dual-map (coordinate map and topological map) and dual-orientation (relative orientation and absolute orientation) strategies, enabling real-time navigation through dynamic scene capture and path planning.<n>Its five core modules-Hippocampal Memory Hub, Visual Cortex Perception Engine, Parietal Spatial Constructor, Prefrontal Decision Center, and Cerebellar Motion Execution Unit-mimic biological cognitive functions to reduce spatial hallucinations and enhance adaptability.
  arXiv  Detail & Related papers  (2025-04-09T02:19:22Z)
• CogNav: Cognitive Process Modeling for Object Goal Navigation with LLMs [39.657663028904814]
  Object goal navigation (ObjectNav) is a fundamental task in embodied AI, requiring an agent to locate a target object in previously unseen environments.<n>We propose CogNav, a framework designed to mimic the cognitive process using large language models.<n>CogNav significantly improves the success rate of ObjectNav at least by relative 14% over the state-of-the-arts.
  arXiv  Detail & Related papers  (2024-12-11T09:50:35Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.