paper-with-me

Papers

Pretrained Transformers Do not Always Improve Robustness

2022-10-14 · Swaroop Mishra, Bhavdeep Singh Sachdeva, Chitta Baral

Pretrained Transformers (PT) have been shown to improve Out of Distribution (OOD) robustness than traditional models such as Bag of Words (BOW), LSTMs, Convolutional Neural Networks (CNN) powered by Word2Vec and Glove embeddings. How does the robustness comparison hold in a real world setting where some part of the dataset can be noisy? Do PT also provide more robust representation than traditional models on exposure to noisy data? We perform a comparative study on 10 models and find an empirical evidence that PT provide less robust representation than traditional models on exposure to noisy data. We investigate further and augment PT with an adversarial filtering (AF) mechanism that has been shown to improve OOD generalization. However, increase in generalization does not necessarily increase robustness, as we find that noisy data fools the AF method powered by PT.

📄 PDF Abstract BibTeX arXiv:2210.07663

Code (0)

등록된 구현이 없습니다.

Methods 이 논문이 사용한 방법론

GloVe GloVe Embeddings are a type of word embedding that encode the co-occurrence probability ratio between two words as vector differences. GloVe uses a weighted least squares…

Similar Papers 제목 키워드 기반

Pretrained Transformers Improve Out-of-Distribution Robustness

2020-04-13 · ACL 2020 6 · Dan Hendrycks, Xiaoyuan Liu, Eric Wallace, Adam Dziedzic 외

Although pretrained Transformers such as BERT achieve high accuracy on in-distribution examples, do they generalize to new distributions? We systematically measure out-of-distribution (OOD) generalization for seven NLP d…

Trade-Off Between Robustness and Accuracy of Vision Transformers

2023-01-01 · CVPR 2023 1 · Yanxi Li, Chang Xu

Although deep neural networks (DNNs) have shown great successes in computer vision tasks, they are vulnerable to perturbations on inputs, and there exists a trade-off between the natural accuracy and robustness to su…

Adversarially Pretrained Transformers may be Universally Robust In-Context Learners

2025-05-20 · Soichiro Kumano, Hiroshi Kera, Toshihiko Yamasaki

Adversarial training is one of the most effective adversarial defenses, but it incurs a high computational cost. In this study, we show that transformers adversarially pretrained on diverse tasks can serve as robust foun…

In-Context Learning

On the Limitations of Sociodemographic Adaptation with Transformers

2022-08-01 · Chia-Chien Hung, Anne Lauscher, Dirk Hovy, Simone Paolo Ponzetto 외

Sociodemographic factors (e.g., gender or age) shape our language. Previous work showed that incorporating specific sociodemographic factors can consistently improve performance for various NLP tasks in traditional NLP m…

Language ModelingLanguage ModellingMulti-Task Learning

Optimal Attention Temperature Improves the Robustness of In-Context Learning under Distribution Shift in High Dimensions

2025-11-03 · Samet Demir, Zafer Dogan arxiv

Pretrained Transformers can perform in-context learning (ICL) from a few demonstrations, but this ability can fail sharply when the test distribution differs from pretraining, a common deployment setting. We study attent…