Communicating Plans, Not Percepts: Scalable Multi-Agent Coordination with Embodied World Models
Fuente:
arXiv
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909921182744576 |
|---|---|
| author | Hill, Brennen A. Wei, Mant Koh En Jishnuanandh, Thangavel |
| author_facet | Hill, Brennen A. Wei, Mant Koh En Jishnuanandh, Thangavel |
| contents | Robust coordination is critical for effective decision-making in multi-agent systems, especially under partial observability. A central question in Multi-Agent Reinforcement Learning (MARL) is whether to engineer communication protocols or learn them end-to-end. We investigate this dichotomy using embodied world models. We propose and compare two communication strategies for a cooperative task-allocation problem. The first, Learned Direct Communication (LDC), learns a protocol end-to-end. The second, Intention Communication, uses an engineered inductive bias: a compact, learned world model, the Imagined Trajectory Generation Module (ITGM), which uses the agent's own policy to simulate future states. A Message Generation Network (MGN) then compresses this plan into a message. We evaluate these approaches on goal-directed interaction in a grid world, a canonical abstraction for embodied AI problems, while scaling environmental complexity. Our experiments reveal that while emergent communication is viable in simple settings, the engineered, world model-based approach shows superior performance, sample efficiency, and scalability as complexity increases. These findings advocate for integrating structured, predictive models into MARL agents to enable active, goal-driven coordination. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_02912 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Communicating Plans, Not Percepts: Scalable Multi-Agent Coordination with Embodied World Models Hill, Brennen A. Wei, Mant Koh En Jishnuanandh, Thangavel Multiagent Systems Artificial Intelligence Machine Learning Systems and Control 68T42, 68T05, 90C40, 93E35, 68T07 I.2.11; I.2.6; I.2.8 Robust coordination is critical for effective decision-making in multi-agent systems, especially under partial observability. A central question in Multi-Agent Reinforcement Learning (MARL) is whether to engineer communication protocols or learn them end-to-end. We investigate this dichotomy using embodied world models. We propose and compare two communication strategies for a cooperative task-allocation problem. The first, Learned Direct Communication (LDC), learns a protocol end-to-end. The second, Intention Communication, uses an engineered inductive bias: a compact, learned world model, the Imagined Trajectory Generation Module (ITGM), which uses the agent's own policy to simulate future states. A Message Generation Network (MGN) then compresses this plan into a message. We evaluate these approaches on goal-directed interaction in a grid world, a canonical abstraction for embodied AI problems, while scaling environmental complexity. Our experiments reveal that while emergent communication is viable in simple settings, the engineered, world model-based approach shows superior performance, sample efficiency, and scalability as complexity increases. These findings advocate for integrating structured, predictive models into MARL agents to enable active, goal-driven coordination. |
| title | Communicating Plans, Not Percepts: Scalable Multi-Agent Coordination with Embodied World Models |
| topic | Multiagent Systems Artificial Intelligence Machine Learning Systems and Control 68T42, 68T05, 90C40, 93E35, 68T07 I.2.11; I.2.6; I.2.8 |
| url | https://arxiv.org/abs/2508.02912 |