VMAS

The 'VMAS' dataset/benchmark is used to evaluate multi-agent reinforcement learning (MARL) algorithms, focusing on their ability to facilitate effective exploration in environments with sparse rewards.

Papers using VMAS (3)

Graph-SND: Sparse Aggregation for Behavioral Diversity in Multi-Agent Reinforcement Learning2026Wonder Wins Ways: Curiosity-Driven Exploration through Multi-Agent Contextual Calibration2025Efficient Model-Based Reinforcement Learning Through Optimistic Thompson Sampling2024 · 1 cites