The Transformer architecture has proven to be highly effective for Autom...
Detecting adversarial samples that are carefully crafted to fool the mod...
Adversarial training is one of the best-performing methods in improving ...
Prompting methods such as Chain-of-Thought (CoT) have shed new light on
...
Pretrained language models have achieved remarkable success in various
n...
Dataset bias, i.e., the over-reliance on dataset-specific literal heuris...
Dataset bias has attracted increasing attention recently for its detrime...
Natural language understanding (NLU) models tend to rely on spurious
cor...