Papers with unsupervised OOD detection method

2 papers
Contrastive Out-of-Distribution Detection for Pretrained Transformers (2021.emnlp-main)

Copied to clipboard

Challenge: Pretrained Transformers achieve remarkable performance when training and test data are from the same distribution, but in real-world scenarios, out-of-distribution instances can cause semantic shift problems.
Approach: They propose to fine-tune the Transformers with a contrastive loss, which improves the compactness of representations, and to use the Mahalanobis distance in the model's penultimate layer to detect OOD instances.
Outcome: The proposed method outperforms baselines in the real-world and achieves near-perfect OOD detection performance.
Semantic Role Labeling Guided Out-of-distribution Detection (2024.lrec-main)

Copied to clipboard

Challenge: Existing methods for identifying domain-shifted instances are prone to OOD and adversarial inputs.
Approach: They propose an unsupervised method that separates, extracts, and learns the semantic role labeling guided out-of-distribution Detection (SRLOOD) they propose a self-supervised approach to enhance global-local feature learning by predicting SRL extracted role.
Outcome: The proposed method achieves SOTA performance on four OOD benchmarks.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations