ETD-MAPPO adds dual-gated epistemic triggers to MAPPO so agents in asynchronous MARL autonomously modulate compute frequency via policy entropy and twin-critic divergence, yielding 73.6% off-ball compute reduction and over 60% performance gains on LBF, MPE, and GRF.
Agent-centric actor-critic for asynchronous multi-agent reinforcement learning
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.MA 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Dual-Gated Epistemic Time-Dilation: Autonomous Compute Modulation in Asynchronous MARL
ETD-MAPPO adds dual-gated epistemic triggers to MAPPO so agents in asynchronous MARL autonomously modulate compute frequency via policy entropy and twin-critic divergence, yielding 73.6% off-ball compute reduction and over 60% performance gains on LBF, MPE, and GRF.