Unsupervised Layer-wise Score Aggregation for Textual OOD Detection

arXiv:2302.09852v3 Announce Type: replace
Abstract: Out-of-distribution (OOD) detection is a rapidly growing field due to new robustness and security requirements driven by an increased number of AI-based systems. Existing OOD textual detectors often rely on an anomaly score (e.g., Mahalanobis distance) computed on the embedding output of the last layer of the encoder. In this work, we observe that OOD detection performance varies greatly depending on the task and layer output. More importantly, we show that the usual choice (the …

abstract aggregation anomaly arxiv cs.ai cs.cl detection distribution embedding layer requirements robustness security systems textual type unsupervised wise