Can we predict how well a team of individuals will perform together?

Understanding semantics of natural language utterances is a fundamental problem in machine learning. Semantics is usually invariant to permute some components in it.

Multi-Agent Reinforcement Learning (MARL) has achieved impressive performance in a wide array of applications including multi-player game play [42, 31], multi-robot systems [13], and autonomousdriving[25].

Using theneural tangent kernel (NTK), we demonstrate that this phenomena occurs inpart because the NTK is not shift-invariant.

Our bound addresses the second question; it suggests that learning algorithms that bias towards models with small variation across the source threat model exhibit smaller drop in robustness to particularunforeseenattacks.

Networked1).LetG=(V,E)beaconnected, thecommunicationE contains (i, j) ifagentsi and j can directlyviamessagest, eachagentj broadcastsmj(t) toalltheirneighbors. times G, after discarded.