Timezone: »

Learning Optimal Representations with the Decodable Information Bottleneck
Yann Dubois · Douwe Kiela · David Schwab · Ramakrishna Vedantam

Tue Dec 08 07:20 AM -- 07:30 AM (PST) @ Orals & Spotlights: Clustering/Ranking

We address the question of characterizing and finding optimal representations for supervised learning. Traditionally, this question has been tackled using the Information Bottleneck, which compresses the inputs while retaining information about the targets, in a decoder-agnostic fashion. In machine learning, however, our goal is not compression but rather generalization, which is intimately linked to the predictive family or decoder of interest (e.g. linear classifier). We propose the Decodable Information Bottleneck (DIB) that considers information retention and compression from the perspective of the desired predictive family. As a result, DIB gives rise to representations that are optimal in terms of expected test performance and can be estimated with guarantees. Empirically, we show that the framework can be used to enforce a small generalization gap on downstream classifiers and to predict the generalization ability of neural networks.

Author Information

Yann Dubois (Facebook AI)
Douwe Kiela (Facebook AI Research)
David Schwab (Facebook AI Research)
Ramakrishna Vedantam (Facebook AI Research)

Related Events (a corresponding poster, oral, or spotlight)

More from the Same Authors