ALFRED

ALFRED is a benchmark dataset that contains tasks for evaluating robotic planning and decision-making capabilities in real-world scenarios.

Papers using ALFRED (7)

On-Device Robotic Planning: Eliminating Inference Redundancy for Efficient Decision-Making2026RePlan-Bot: Multi-Level Replanning for Embodied Instruction Following2026FlowPlan: Zero-Shot Task Planning with LLM Flow Engineering for Robotic Instruction Following2025 · 1 citesThink Twice, Act Once: Verifier-Guided Action Selection For Embodied Agents2026ESCAPE: Episodic Spatial Memory and Adaptive Execution Policy for Long-Horizon Mobile Manipulation2026When Should a Robot Think? Resource-Aware Reasoning via Reinforcement Learning for Embodied Robotic Decision-Making2026KEEP: A KV-Cache-Centric Memory Management System for Efficient Embodied Planning2026