AI Paper Atlas

Research map / Multi-agent systems and games / Cooperative multi-agent RL

ADAPT: Adaptive Decentralized Architecture with Perception-Aligned Training for Structural Generalization in Multi-Agent RL

Zhixiang Zhang, Shuo Chen, Yexin Li, Feng Wang

AAAI 2026 · Multi-agent systems and games › Cooperative multi-agent RL · approach: Cooperative MARL algorithms · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
select actions with FINE instead of majority voting among source policies (ablation)ADAPT · SMAC zero-shot 5m to 7mtest win rate (%): 73.44 -> 85.76Improved
select actions with FINE instead of majority voting among source policies (ablation)ADAPT · SMAC zero-shot 3m to 5mtest win rate (%): 43.13 -> 47.16Improved
select actions with FINE instead of voting (ablation)ADAPT · MPE zero-shot (12 target maps)test win rate (%): better on 9 of 12 mapsMixed
select actions with FINE instead of voting (ablation)ADAPT · MPE 3prey-3predator to 12prey-10predatortest win rate (%): 53.17 -> 73.50Improved
select actions with FINE instead of voting (ablation)ADAPT · MPE 2prey-4predator to 6prey-4predatortest win rate (%): 84.90 -> 75.60Hurt
use ADAPT instead of I2Q, trained from scratchI2Q · SMAC 5m_vs_6mtest win rate (%): 40.00 -> 86.25Improved
use ADAPT instead of I2Q, trained from scratchI2Q · SMAC 2s3zdead allies: 2.89 -> 1.33Improved
use ADAPT instead of I2Q, trained from scratchI2Q · SMAC 5mtest win rate (%): 99.38 -> 98.13No clear effect
use ADAPT+FINE zero-shot instead of OPT-DTDEOPT-DTDE · SMAC zero-shot 5m to 7mtest win rate (%): 20.20 -> 85.76Improved
use ADAPT-share (shared gradients, identical networks) instead of independent networks with FINEADAPT · MPE zero-shot 3prey-3predator to 20prey-20predatortest win rate (%): 79.17 -> 63.50Hurt
ADAPT+FINE vs OPT-CTDE zero-shot after training on 2prey-4predatorADAPT+FINE · MPE 2prey-4predator sourcetest win rate (%): exceeds OPT-CTDE on 4 of 6 zero-shot mapsMixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy