Sequence Modeling for N-Agent Ad Hoc Teamwork
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_ | 1866914102180315136 |
|---|---|
| author | Wang, Caroline Shi, Di Yang Liebman, Elad Durugkar, Ishan Rahman, Arrasy Stone, Peter |
| author_facet | Wang, Caroline Shi, Di Yang Liebman, Elad Durugkar, Ishan Rahman, Arrasy Stone, Peter |
| contents | N-agent ad hoc teamwork (NAHT) is a newly introduced challenge in multi-agent reinforcement learning, where controlled subteams of varying sizes must dynamically collaborate with varying numbers and types of unknown teammates without pre-coordination. The existing learning algorithm (POAM) considers only independent learning for its flexibility in dealing with a changing number of agents. However, independent learning fails to fully capture the inter-agent dynamics essential for effective collaboration. Based on our observation that transformers deal effectively with sequences with varying lengths and have been shown to be highly effective for a variety of machine learning problems, this work introduces a centralized, transformer-based method for N-agent ad hoc teamwork. Our proposed approach incorporates historical observations and actions of all controlled agents, enabling optimal responses to diverse and unseen teammates in partially observable environments. Empirical evaluation on a StarCraft II task demonstrates that MAT-NAHT outperforms POAM, achieving superior sample efficiency and generalization, without auxiliary agent-modeling objectives. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_05527 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Sequence Modeling for N-Agent Ad Hoc Teamwork Wang, Caroline Shi, Di Yang Liebman, Elad Durugkar, Ishan Rahman, Arrasy Stone, Peter Multiagent Systems N-agent ad hoc teamwork (NAHT) is a newly introduced challenge in multi-agent reinforcement learning, where controlled subteams of varying sizes must dynamically collaborate with varying numbers and types of unknown teammates without pre-coordination. The existing learning algorithm (POAM) considers only independent learning for its flexibility in dealing with a changing number of agents. However, independent learning fails to fully capture the inter-agent dynamics essential for effective collaboration. Based on our observation that transformers deal effectively with sequences with varying lengths and have been shown to be highly effective for a variety of machine learning problems, this work introduces a centralized, transformer-based method for N-agent ad hoc teamwork. Our proposed approach incorporates historical observations and actions of all controlled agents, enabling optimal responses to diverse and unseen teammates in partially observable environments. Empirical evaluation on a StarCraft II task demonstrates that MAT-NAHT outperforms POAM, achieving superior sample efficiency and generalization, without auxiliary agent-modeling objectives. |
| title | Sequence Modeling for N-Agent Ad Hoc Teamwork |
| topic | Multiagent Systems |
| url | https://arxiv.org/abs/2506.05527 |