Inducing Functions through Reinforcement Learning without Task Specification

TL;DR

This paper presents a bio-inspired reinforcement learning framework that enables neural networks to develop high-level functions like image classification and hidden variable estimation without explicit task training or pre-specification.

Contribution

It introduces a novel approach where neural networks induce complex functions through environmental interaction, inspired by biological cognitive development.

Findings

High-level functions can be induced without pre-training.

Neural networks develop functions like image classification naturally.

Simultaneous induction of multiple functions observed.

Abstract

We report a bio-inspired framework for training a neural network through reinforcement learning to induce high level functions within the network. Based on the interpretation that animals have gained their cognitive functions such as object recognition - without ever being specifically trained for - as a result of maximizing their fitness to the environment, we place our agent in an environment where developing certain functions may facilitate decision making. The experimental results show that high level functions, such as image classification and hidden variable estimation, can be naturally and simultaneously induced without any pre-training or specifying them.