Dynamically handling task disruptions by composing together behavior modules

TL;DR

This paper explores how neural networks can dynamically compose behavior modules to adapt to task disruptions, inspired by biological neural systems, and evaluates different architectures for this capability.

Contribution

It introduces a method for dynamically composing path modules to handle task disruptions and compares multiple neural network architectures for this purpose.

Findings

LSTM and Morphognosis outperform other models in dynamic path composition.

The approach enables neural networks to adapt to novel, piecemeal learned paths.

Network performance improves with architectures capable of handling sequential and hierarchical information.

Abstract

Biological neural networks operate in the presence of task disruptions as they guide organisms toward goals. A familiar stream of stimulus-response causations can be disrupted by subtask streams imposed by the environment. For example, taking a familiar path to a foraging area might be disrupted by the presence of a predator, necessitating a "detour" to the area. The detour can be a known alternative path that must be dynamically composed with the original path to accomplish the overall task. In this project, overarching base paths are disrupted by independently learned path modules in the form of insertion, substitution, and deletion modifications to the base paths such that the resulting modified paths are novel to the network. The network's performance is then tested on these paths that have been learned in piecemeal fashion. In sum, the network must compose a new task on the fly. Several network architectures are tested: Time delay neural network (TDNN), Long short-term memory (LSTM), Temporal convolutional network (TCN), and Morphognosis, a hierarchical neural network. LSTM and Morphognosis perform significantly better for this task.