$k$Folden: $k$-Fold Ensemble for Out-Of-Distribution Detection

TL;DR

kFolden is a novel ensemble framework for out-of-distribution detection in NLP that trains sub-models on masked categories to improve OOD detection without external data, outperforming existing methods.

Contribution

The paper introduces kFolden, a simple ensemble approach that simulates OOD scenarios during training, enhancing detection performance without external datasets.

Findings

kFolden outperforms existing OOD detection methods in benchmarks.

It maintains high in-domain classification accuracy.

The framework is effective across multiple text classification datasets.

Abstract

Out-of-Distribution (OOD) detection is an important problem in natural language processing (NLP). In this work, we propose a simple yet effective framework $k$Folden, which mimics the behaviors of OOD detection during training without the use of any external data. For a task with $k$ training labels, $k$Folden induces $k$ sub-models, each of which is trained on a subset with $k-1$ categories with the left category masked unknown to the sub-model. Exposing an unknown label to the sub-model during training, the model is encouraged to learn to equally attribute the probability to the seen $k-1$ labels for the unknown label, enabling this framework to simultaneously resolve in- and out-distribution examples in a natural way via OOD simulations. Taking text classification as an archetype, we develop benchmarks for OOD detection using existing text classification datasets. By conducting comprehensive comparisons and analyses on the developed benchmarks, we demonstrate the superiority of $k$Folden against current methods in terms of improving OOD detection performances while maintaining improved in-domain classification accuracy. The code and datasets can be found at: \url{https://github.com/ShannonAI/kfolden-ood-detection}.