Drop-Bottleneck: Learning Discrete Compressed Representation for Noise-Robust Exploration
This addresses the challenge of exploration in noisy, reward-sparse environments for reinforcement learning agents, offering a new framework with specific improvements over existing methods.
The paper tackles the problem of learning discrete compressed representations for noise-robust exploration in reinforcement learning by proposing Drop-Bottleneck, a novel information bottleneck method that drops irrelevant features. It achieves state-of-the-art performance in noisy maze navigation tasks and outperforms Variational Information Bottleneck in aspects like adversarial robustness.
We propose a novel information bottleneck (IB) method named Drop-Bottleneck, which discretely drops features that are irrelevant to the target variable. Drop-Bottleneck not only enjoys a simple and tractable compression objective but also additionally provides a deterministic compressed representation of the input variable, which is useful for inference tasks that require consistent representation. Moreover, it can jointly learn a feature extractor and select features considering each feature dimension's relevance to the target task, which is unattainable by most neural network-based IB methods. We propose an exploration method based on Drop-Bottleneck for reinforcement learning tasks. In a multitude of noisy and reward sparse maze navigation tasks in VizDoom (Kempka et al., 2016) and DMLab (Beattie et al., 2016), our exploration method achieves state-of-the-art performance. As a new IB framework, we demonstrate that Drop-Bottleneck outperforms Variational Information Bottleneck (VIB) (Alemi et al., 2017) in multiple aspects including adversarial robustness and dimensionality reduction.