AdvGLUE (Adversarial GLUE)

Adversarial restatement of five GLUE tasks using word-level, sentence-level and human-written attacks; OpenCompass scores a public annotated development pack, not the original hidden test set.

Also known as: Adversarial GLUE, AdvGLUE

unassessed

This page is a discovery lead. Nobody has yet assessed it against the catalogue contract, so it carries no disposition. Absence of evidence here is not evidence of staleness.
Categorycomposite
Subcategoryadversarial robustness suite over five GLUE tasks (SST-2, QQP, QNLI, RTE, MNLI)
Page statusactive
Metricper-task GLUE metrics averaged as an AdvGLUE score; OpenCompass additionally reports acc_drop = 1 - (adversarial accuracy / original accuracy)
Directionhigher_is_better
Dataset size4978
Dataset licenceCC-BY-SA-4.0
PublisherUniversity of Illinois Urbana-Champaign and Microsoft

What it measures

AdvGLUE asks whether a model that looks strong on GLUE still labels the same sentence-pair and sentiment tasks after adversarial edits. It covers five GLUE tasks: SST-2 sentiment, QQP paraphrase, and the NLI tasks MNLI, QNLI and RTE. CoLA and WNLI from GLUE are omitted. Attacks include 14 methods grouped as word-level (embedding similarity, typos, context-aware, knowledge-guided, compositions), sentence-level (syntactic and distraction), and human-crafted sets (CheckList, StressTest, ANLI, AdvSQuAD). Crowd workers filtered examples so that the label should still hold.

Task format

Same classification or NLI format as the parent GLUE task, on adversarially perturbed text. OpenCompass instead prompts for an option letter and reports accuracy drop against concatenated original items.

Models reporting this benchmark

No model card in ModelSpec reports this benchmark yet.

Data

This page as JSON · Edit on GitHub