Natural Language Processing (NLP) models optimized for predictive performance often make high confidence errors and suffer from vulnerability to adversarial and out-of-distribution data. Existing work has mainly focused on mitigation of such errors using either humans or an automated approach. In this study, we explore the usage of large language models (LLMs) for data augmentation as a potential solution to the issue of NLP models making wrong predictions with high confidence during classification tasks. We compare the effectiveness of synthetic data generated by LLMs with that of human data obtained via the same procedure. For mitigation, humans or LLMs provide natural language characterizations of high confidence misclassifications to generate synthetic data, which are then used to extend the training set. We conduct an extensive evaluation of our approach on three classification tasks and demonstrate its effectiveness in reducing the number of high confidence misclassifications present in the model, all while maintaining the same level of accuracy. Moreover, we find that the cost gap between humans and LLMs surpasses an order of magnitude, as LLMs attain human-like performance while being more scalable.

Assessing the predictive accuracy of black box classifiers is challenging in the absence of labeled test datasets. In these scenarios we may need to rely on a human oracle to evaluate individual predictions; presenting the challenge to create query algorithms to guide the search for points that provide the most information about the classifier's predictive characteristics. Previous works have focused on developing utility models and query algorithms for discovering unknown unknowns --- misclassifications with a predictive confidence above some arbitrary threshold. However, if misclassifications occur at the rate reflected by the confidence values, then these search methods reveal nothing more than a proper assessment of predictive certainty. We are unable to properly mitigate the risks associated with model deficiency when the model's confidence in prediction exceeds the actual model accuracy. We propose a facility locations utility model and corresponding greedy query algorithm that instead searches for overconfident unknown unknowns. Through robust empirical experiments we demonstrate that the greedy query algorithm with the facility locations utility model consistently results in oracle queries with superior performance in discovering overconfident unknown unknowns than previous methods.

A classifier’s low confidence in prediction is often indicative of whether its prediction will be wrong; in this case, inputs are called known unknowns. In contrast, unknown unknowns (UUs) are inputs on which a classifier makes a high confidence mistake. Identifying UUs is especially important in safety-critical domains like medicine (diagnosis) and law (recidivism prediction). Previous work by Lakkaraju et al. (2017) on identifying unknown unknowns assumes that the utility of each revealed UU is independent of the others, rather than considering the set holistically. While this assumption yields an efficient discovery algorithm, we argue that it produces an incomplete understanding of the classifier’s limitations. In response, this paper proposes a new class of utility models that rewards how well the discovered UUs cover (or "explain") a sample distribution of expected queries. Although choosing an optimal cover is intractable, even if the UUs were known, our utility model is monotone submodular, affording a greedy discovery strategy. Experimental results on four datasets show that our method outperforms bandit-based approaches and achieves within 60.9% utility of an omniscient, tractable upper bound.