ALFWorld

ALFWorld is a benchmark dataset that contains multimodal environments used to evaluate the performance of embodied agents in perception and decision-making tasks.

Papers using ALFWorld (10)

PRISM: Perception Reasoning Interleaved for Sequential Decision Making2026MemCompiler: Compile, Don't Inject -- State-Conditioned Memory for Embodied Agents2026NeSyPr: Neurosymbolic Proceduralization For Efficient Embodied Reasoning2025Constrained Natural Language Action Planning for Resilient Embodied Systems2025MCTS-EP: Empowering Embodied Planning with Online Preference Optimization2025World Model Implanting for Test-time Adaptation of Embodied Agents2025FCRF: Flexible Constructivism Reflection for Long-Horizon Robotic Task Planning with Large Language Models2025Unleashing Embodied Task Planning Ability in LLMs via Reinforcement Learning2025SEEA-R1: Tree-Structured Reinforcement Fine-Tuning for Self-Evolving Embodied Agents2025NeSyC: A Neuro-symbolic Continual Learner For Complex Embodied Tasks In Open Domains2025