MiniGrid
Emerging11papers using it
2021first seen
MiniGrid is a benchmark that contains a set of gridworld environments used to evaluate the performance of embodied agents in tasks involving navigation and interaction.
Papers using MiniGrid (9)
- Adaptive Context Length Optimization with Low-Frequency Truncation for Multi-Agent Reinforcement LearningAgentSpec: Understanding Embodied Agent Scaffolds Through Controlled CompositionWorld Action Verifier: Self-Improving World Models via Forward-Inverse AsymmetryBeyond Fixed Tasks: Unsupervised Environment Design for Task-Level PairsSoftware Engineering Agents For Embodied Controller Generation : A Study In Minigrid EnvironmentsCode-driven Planning In Grid Worlds With Large Language ModelsA representational framework for learning and encoding structurally enriched trajectories in complex agent environmentsSuccessor Feature Landmarks for Long-Horizon Goal-Conditioned
Reinforcement LearningLLM Augmented Hierarchical Agents