Grounding from an AI and Cognitive Science Lens

TL;DR

The article examines the potential of neurosymbolic approaches tailored for grounding tasks, showcasing how they can more comprehensively address grounding.

Abstract

Grounding is a challenging problem, requiring a formal definition and different levels of abstraction. This article explores grounding from both cognitive science and machine learning perspectives. It identifies the subtleties of grounding, its significance for collaborative agents, and similarities and differences in grounding approaches in both communities. The article examines the potential of neuro-symbolic approaches tailored for grounding tasks, showcasing how they can more comprehensively address grounding. Finally, we discuss areas for further exploration and development in grounding.

Figures (2)

• Figure 1: Types of Grounding
• Figure 2: Grounding has multiple dimensions and needs knowledge at different levels of abstractions (this is similar to the need for linguistic, common sense, world model, and domain knowledge for language understanding; see Fig. 3 of (https://bit.ly/KiLU). Grounding may occur at different levels in the task execution process. For example, the drone agent must be able to comprehend the instructions in a natural language format. Next, the instruction may require parsing relevant symbols (in the case of neuro-symbolic methods) for reasoning processes. The information extracted must then be grounded to the drone's capabilities (common-sense grounding). Finally, the instruction must be grounded in the specific navigation task.