no code implementations • Findings (EMNLP) 2021 • Chang Xu, Jun Wang, Francisco Guzmán, Benjamin Rubinstein, Trevor Cohn
NLP models are vulnerable to data poisoning attacks.
no code implementations • ACL 2022 • Jun Wang, Benjamin Rubinstein, Trevor Cohn
In this paper we describe a new source of bias prevalent in NMT systems, relating to translations of sentences containing person names.
1 code implementation • 25 May 2023 • Xuanli He, Jun Wang, Benjamin Rubinstein, Trevor Cohn
Backdoor attacks are an insidious security threat against machine learning models.
1 code implementation • 19 May 2023 • Xuanli He, Qiongkai Xu, Jun Wang, Benjamin Rubinstein, Trevor Cohn
Modern NLP models are often trained over large untrusted datasets, raising the potential for a malicious adversary to compromise model behaviour.
1 code implementation • NeurIPS 2021 • Zhuolin Yang, Linyi Li, Xiaojun Xu, Shiliang Zuo, Qian Chen, Benjamin Rubinstein, Pan Zhou, Ce Zhang, Bo Li
To answer these questions, in this work we first theoretically analyze and outline sufficient conditions for adversarial transferability between models; then propose a practical algorithm to reduce the transferability between base models within an ensemble to improve its robustness.
no code implementations • 22 Dec 2015 • Zuhe Zhang, Benjamin Rubinstein, Christos Dimitrakakis
We study how to communicate findings of Bayesian inference to third parties, while preserving the strong guarantee of differential privacy.
no code implementations • 5 Jun 2013 • Christos Dimitrakakis, Blaine Nelson, and Zuhe Zhang, Aikaterini Mitrokotsa, Benjamin Rubinstein
All our general results hold for arbitrary database metrics, including those for the common definition of differential privacy.