iQRL -- Implicitly Quantized Representations for Sample-efficient Reinforcement Learning

TL;DR

iQRL introduces an efficient, quantized representation learning approach for reinforcement learning that enhances sample efficiency and outperforms existing methods in continuous control tasks.

Contribution

The paper presents a novel quantization technique for latent representations in RL, improving stability and performance without complex additional components.

Findings

Outperforms recent representation learning methods in DeepMind Control Suite benchmarks.

Prevents representation collapse through latent state quantization.

Compatible with any model-free RL algorithm.

Abstract

Learning representations for reinforcement learning (RL) has shown much promise for continuous control. We propose an efficient representation learning method using only a self-supervised latent-state consistency loss. Our approach employs an encoder and a dynamics model to map observations to latent states and predict future latent states, respectively. We achieve high performance and prevent representation collapse by quantizing the latent representation such that the rank of the representation is empirically preserved. Our method, named iQRL: implicitly Quantized Reinforcement Learning, is straightforward, compatible with any model-free RL algorithm, and demonstrates excellent performance by outperforming other recently proposed representation learning methods in continuous control benchmarks from DeepMind Control Suite.