Contrastive Novelty-Augmented Learning: Anticipating Outliers with Large Language Models

In many task settings, text classification models are likely to encounter examples from novel classes on which they cannot predict correctly. Selective prediction, in which models abstain on low-confidence examples, provides a possible solution, but existing models are often overly confident on unseen classes. To remedy this overconfidence, we introduce Contrastive Novelty-Augmented Learning (CoNAL), a twostep method that generates OOD examples representative of novel classes, then trains to decrease confidence on them. First, we generate OOD examples by prompting a large language model twice: we prompt it to enumerate relevant novel classes, then generate examples from each novel class matching the task format. Second, we train a classifier with a novel contrastive objective that encourages lower confidence on generated OOD examples than training examples. When trained with CoNAL, classifiers improve in their ability to detect and abstain on novel class examples over prior methods by an average of 2.3% in terms of accuracy under the accuracy-coverage curve (AUAC) and 5.5% AUROC across 4 NLP datasets, with no cost to in-distribution accuracy. 1 Novelty Prompting CCL Training Selective Prediction ✅ Yankees Unlikely to Get D-Rays… A new report from the Canadian Food… Sports World Business Sports World Business Kyrie Irving returns to play in… The Cartwheel Galaxy Is the… " Generate a diverse list of news categories: world, sports, business, ID Labels science, crime, travel, auto, food OOD Labels Given a label, generate a corresponding example: world Bush, Kerry Trade Barbs Following… sports Yankees Unlikely to Get D-Rays… business ECB sees gradual recovery in eurozone… Example Generation Generation Label Generation