From Eye-blinks to State Construction: Diagnostic Benchmarks for Online Representation Learning

TL;DR

This paper introduces three diagnostic benchmarks inspired by classical conditioning to evaluate and improve online prediction learning in recurrent neural networks, emphasizing state construction and temporal association capabilities.

Contribution

The paper proposes new diagnostic problems for online prediction, inspired by classical conditioning, to assess and advance recurrent neural network learning methods in continual settings.

Findings

Current training methods are expensive for online prediction.

The proposed problems are challenging but suitable for small-scale testing.

They highlight limitations and guide improvements in online representation learning.

Abstract

We present three new diagnostic prediction problems inspired by classical-conditioning experiments to facilitate research in online prediction learning. Experiments in classical conditioning show that animals such as rabbits, pigeons, and dogs can make long temporal associations that enable multi-step prediction. To replicate this remarkable ability, an agent must construct an internal state representation that summarizes its interaction history. Recurrent neural networks can automatically construct state and learn temporal associations. However, the current training methods are prohibitively expensive for online prediction -- continual learning on every time step -- which is the focus of this paper. Our proposed problems test the learning capabilities that animals readily exhibit and highlight the limitations of the current recurrent learning methods. While the proposed problems are nontrivial, they are still amenable to extensive testing and analysis in the small-compute regime, thereby enabling researchers to study issues in isolation, ultimately accelerating progress towards scalable online representation learning methods.