D
Dawn Song
Researcher at University of California, Berkeley
Publications - 504
Citations - 75245
Dawn Song is an academic researcher from University of California, Berkeley. The author has contributed to research in topics: Computer science & Deep learning. The author has an hindex of 117, co-authored 460 publications receiving 61572 citations. Previous affiliations of Dawn Song include FireEye, Inc. & University of California.
Papers
More filters
Beyond Output Voting: Detecting Compromised Replicas using Behavioral Distance
TL;DR: This paper presents a novel architecture to detect mimicry attacks using “behavioral distance”, by which two diverse replicas processing the same inputs are continually monitored to detect divergence in their low-level behaviors and hence potentially the compromise of one of them.
Posted Content
PrivFramework: A System for Configurable and Automated Privacy Policy Compliance.
TL;DR: PrivFramework allows data owners to write powerful privacy policies to protect their data and automatically enforces these policies against analysis programs written in Python.
On a First Step to the Automatic Generation of Security Protocols
Adrian Perrig,Dawn Song +1 more
TL;DR: Automatic protocol generation (APG for short), a novel mechanism to generate security protocols automatically which has minimal cost with respect to the metric function, as well as satisfies the security properties and system requirements.
Journal ArticleDOI
DecodingTrust: A Comprehensive Assessment of Trustworthiness in GPT Models
Boxin Wang,Weixin Chen,Hengzhi Pei,Chulin Xie,Mintong Kang,Chenhui Zhang,Chejian Xu,Zidi Xiong,Ritik Dutta,Rylan Schaeffer,Sang Truong,Mantas Mazeika,Dan Hendrycks,Zi-Han Lin,Yuk-Kit Cheng,Sanmi Koyejo,Dawn Song,Bo Li +17 more
TL;DR: This article proposed a comprehensive trustworthiness evaluation for large language models with a focus on Generative Pre-trained Transformer (GPT) models and found that GPT models can be easily misled to generate toxic and biased outputs and leak private information in both training data and conversation history.
Journal ArticleDOI
Multi-Factor Key Derivation Function (MFKDF)
Vivek Nair,Dawn Song +1 more
TL;DR: This work presents the first general construction of a Multi-Factor Key Derivation Function (MFKDF), and demonstrates the ability of this function to not only improve the security of existing systems implementing PBKDFs, but also to enable new applications where PBK DFs would not be considered a feasible approach.