Analyzing And Editing Inner Mechanisms of Backdoored Language Models

Author:

Lamparth Max1ORCID,Reuel Anka1ORCID

Affiliation:

1. Stanford University, USA

Publisher

ACM

Reference66 articles.

1. Gorka Abad, Servio Paguada, Oğuzhan Ersoy, Stjepan Picek, Víctor Julio Ramírez-Durán, and Aitor Urbieta. 2023. Sniper Backdoor: Single Client Targeted Backdoor Attack in Federated Learning. In 2023 IEEE Conference on Secure and Trustworthy Machine Learning (SaTML) (Raleigh, NC, USA, 2023-02). IEEE, 377–391.

2. Hojjat Aghakhani, Lea Schönherr, Thorsten Eisenhofer, Dorothea Kolossa, Thorsten Holz, Christopher Kruegel, and Giovanni Vigna. 2023. Venomave: Targeted Poisoning Against Speech Recognition. In 2023 IEEE Conference on Secure and Trustworthy Machine Learning (SaTML) (Raleigh, NC, USA, 2023-02). IEEE, 404–417.

3. Optuna

4. Eugene Bagdasaryan and Vitaly Shmatikov. 2021. Spinning Sequence-to-Sequence Models with Meta-Backdoors. arXiv preprint arXiv:2107.10443 (2021).

5. Clark Barrett, Brad Boyd, Ellie Burzstein, Nicholas Carlini, Brad Chen, Jihye Choi, Amrita Roy Chowdhury, Mihai Christodorescu, Anupam Datta, Soheil Feizi, Kathleen Fisher, Tatsunori Hashimoto, Dan Hendrycks, Somesh Jha, Daniel Kang, Florian Kerschbaum, Eric Mitchell, John Mitchell, Zulfikar Ramzan, Khawaja Shams, Dawn Song, Ankur Taly, and Diyi Yang. 2023. Identifying and Mitigating the Security Risks of Generative AI. arXiv preprint arXiv:2308.14840 (2023).

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3