DR. WELL: Dynamic Reasoning and Learning with Symbolic World Model for Embodied LLM-Based Multi-Agent Collaboration

2025-11-06

AI Summary

Addressing challenges in multi-agent collaborative planning—including partial observability, communication constraints, and trajectory-level coordination failures (e.g., conflicts arising from temporal or kinematic misalignment)—this paper proposes a decentralized neuro-symbolic framework. The method integrates large language models with symbolic reasoning to construct a dynamically updated symbolic world model, and introduces a two-stage role negotiation mechanism that enables agents to reach semantic consensus in a high-level action space, thereby generating interpretable and reusable joint plans. Each agent then executes its assigned actions independently, with behavior grounded in the shared plan. Evaluated on a collaborative block-pushing task, the approach significantly improves task success rate and execution efficiency. Results demonstrate that dynamic learning coupled with decentralized negotiation is essential for sustaining and evolving collaborative capabilities over time.

Technology Category

• Multiagent Systems: Multiagent Planning
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Machine Learning: Neuro-Symbolic Learning

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Search and Retrieval-Augmented AI: Agentic search
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Cooperative multi-agent planning requires agents to make joint decisions with partial information and limited communication. Coordination at the trajectory level often fails, as small deviations in timing or movement cascade into conflicts. Symbolic planning mitigates this challenge by raising the level of abstraction and providing a minimal vocabulary of actions that enable synchronization and collective progress. We present DR. WELL, a decentralized neurosymbolic framework for cooperative multi-agent planning. Cooperation unfolds through a two-phase negotiation protocol: agents first propose candidate roles with reasoning and then commit to a joint allocation under consensus and environment constraints. After commitment, each agent independently generates and executes a symbolic plan for its role without revealing detailed trajectories. Plans are grounded in execution outcomes via a shared world model that encodes the current state and is updated as agents act. By reasoning over symbolic plans rather than raw trajectories, DR. WELL avoids brittle step-level alignment and enables higher-level operations that are reusable, synchronizable, and interpretable. Experiments on cooperative block-push tasks show that agents adapt across episodes, with the dynamic world model capturing reusable patterns and improving task completion rates and efficiency. Experiments on cooperative block-push tasks show that our dynamic world model improves task completion and efficiency through negotiation and self-refinement, trading a time overhead for evolving, more efficient collaboration strategies.

Problem

• Addresses multi-agent coordination failures from trajectory deviations
• Enables symbolic planning abstraction for reusable synchronized operations
• Improves task completion through dynamic world model negotiation

Innovation

• Decentralized neurosymbolic framework for multi-agent planning
• Two-phase negotiation protocol for role allocation
• Symbolic world model enabling reusable synchronized operations