Awesome World Model Hub Papers · Datasets · Projects
← Back to papers

Deductive Chain-of-Thought Augmented Socially-aware Robot Navigation World Model

arXiv 25.10 2025 50.8 method

TLDR

NaviWM augments LLM reasoning with a spatial-temporal world model and deductive chain-of-thought for socially-aware robot navigation, improving success and reducing violations.

Reasoning

The paper introduces a novel integration of formal logic with LLMs for interpretable reasoning in navigation, showing empirical improvements. However, it lacks explicit real-world validation and may face scalability issues with complex social norms.

Read-first score

Read-first score 50.8, weighted from topical fit, citation, graph, method, reproducibility, and recency signals. Original total remains 32.

Recency 8%
86.7

Uses a gentle age decay so recent papers surface without erasing older foundations. 2025

Methodology quality 25%
60

Screens visible abstract and analysis fields for experiment, dataset, baseline, metric, and limitation evidence. markers=experiment,result

Topical relevance 42%
45.7

Uses existing LLM keyword relevance scores normalized to 0-100. world model,world simulator,generative world model,interactive world model,video world model,world dynamics prediction,model-based reinforcement learning world model

Reproducibility 25%
38

Screens links and visible text for paper, code, dataset, artifact, and repository signals. pdf=True; code=False; dataset=False; markers=code

Field roles

Frontier

Rank sensitivity

Stability: volatile; rank range: 352.

Keyword Scores

world model
10
world dynamics prediction
8
interactive world model
6
world simulator
3
generative world model
2
model-based reinforcement learning world model
2
video world model
1

Deep Analysis

Innovations

  • Integration of a spatial-temporal world model with a deductive reasoning module for socially-aware robot navigation
  • Encoding social norms as first-order logic for interpretable and verifiable reasoning
  • Logic-driven chain-of-thought process that guides LLMs through multi-step inference for navigation decisions

Methodology

NaviWM consists of a spatial-temporal world model that captures positions, velocities, and activities of agents, and a deductive reasoning module that guides LLMs through a multi-step, logic-based inference process. Social norms are encoded as first-order logic to enable interpretable and verifiable reasoning. The system is evaluated in crowded environments, comparing against previous prompting or fine-tuning methods.

Key Results

Experiments show that NaviWM improves success rates and reduces social violations, particularly in crowded environments.

Tags