ALFRED

ALFRED is a benchmark dataset that contains tasks for embodied agents to evaluate their ability to understand and execute instructions in a simulated environment.

Papers using ALFRED (9)

One Step At A Time: Long-horizon Vision-and-language Navigation With Milestones2022 · 21 citesReactree: Hierarchical LLM Agent Trees With Control Flow For Long-horizon Task Planning2025 · 1 citesAgentSpec: Understanding Embodied Agent Scaffolds Through Controlled Composition2026When Should a Robot Think? Resource-Aware Reasoning via Reinforcement Learning for Embodied Robotic Decision-Making2026Grounding Language Models with Semantic Digital Twins for Robotic Planning2025Tinyllm: Evaluation And Optimization Of Small Language Models For Agentic Tasks On Edge Devices2025A Persistent Spatial Semantic Representation for High-level Natural Language Instruction Execution2021 · 30 citesEPO: Hierarchical LLM Agents with Environment Preference Optimization2024 · 3 citesLoTa-Bench: Benchmarking Language-oriented Task Planners for Embodied Agents2024