ALFWorld: Aligning Text and Embodied Environments for Interactive Learning

TL;DR

ALFWorld is a simulator that bridges abstract text-based reasoning and concrete visual environment execution, enabling embodied agents to learn and generalize better across tasks.

Contribution

The paper introduces ALFWorld, a novel simulator that aligns abstract text policies with visual environment actions, improving agent generalization in embodied AI tasks.

Findings

Agents trained with ALFWorld generalize better to visual tasks.

The modular design facilitates focused improvements in reasoning, planning, and perception.

Empirical results show improved performance over training solely in visual environments.

Abstract

Given a simple request like Put a washed apple in the kitchen fridge, humans can reason in purely abstract terms by imagining action sequences and scoring their likelihood of success, prototypicality, and efficiency, all without moving a muscle. Once we see the kitchen in question, we can update our abstract plans to fit the scene. Embodied agents require the same abilities, but existing work does not yet provide the infrastructure necessary for both reasoning abstractly and executing concretely. We address this limitation by introducing ALFWorld, a simulator that enables agents to learn abstract, text based policies in TextWorld (C\^ot\'e et al., 2018) and then execute goals from the ALFRED benchmark (Shridhar et al., 2020) in a rich visual environment. ALFWorld enables the creation of a new BUTLER agent whose abstract knowledge, learned in TextWorld, corresponds directly to concrete, visually grounded actions. In turn, as we demonstrate empirically, this fosters better agent generalization than training only in the visually grounded environment. BUTLER's simple, modular design factors the problem to allow researchers to focus on models for improving every piece of the pipeline (language understanding, planning, navigation, and visual scene understanding).