Probing Dec-POMDP Reasoning in Cooperative MARL Conference: Proc. of the 25th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2026); May 25 – 29, 2026; Paphos, Cyprus 1235

Kale-ab Abebe Tessera, Leonard Hinckeldey, Riccardo Zamboni, David Abel, Amos Storkey · 2026

Read Probing Dec-POMDP Reasoning in Cooperative MARL Conference: Proc. of the 25th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2026); May 25 – 29, 2026; Paphos, Cyprus 1235 by Kale-ab Abebe Tessera, Leonard Hinckeldey, Riccardo Zamboni, David Abel, Amos Storkey (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-20804.txt.

Description

Kale-ab Tessera and four co-authors ask whether popular cooperative game benchmarks truly test the reasoning they claim to test. In a decentralized partially observable setting, agents must infer hidden states from history and coordinate on local information alone, a demanding pair of skills.

Their diagnostic suite audits two standard baselines, IPPO and MAPPO, across 37 scenarios spanning five game families. Reactive policies match memory-based agents in over half the cases, and coordination often rests on brittle synchronous action coupling rather than lasting temporal influence, suggesting current benchmarks flatter progress.

Catalog record

CREATORKale-ab Abebe Tessera, Leonard Hinckeldey, Riccardo Zamboni, David Abel, Amos Storkey
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE80 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602-20804

Browse the full library · About the Tower of Babel