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
StackPi: A New Defense Mechanism against IP Spoofing and DDoS Attacks
TL;DR: StackPi allows the host being attacked, or its upstream ISP, to filter out attack packets and to detect spoofed source IP addresses, on a per-packet basis, and can defend against not only DDoS attacks, but also many IP spoofing attacks -- such as TCP hijacking, and multicast source spoofing attack.
Proceedings Article
Improving Neural Program Synthesis with Inferred Execution Traces
TL;DR: This work splits the process into two parts: infer the trace from the input/output example, then infer the program from the trace, which leads to state-of-the-art results in program synthesis in the Karel domain.
Book ChapterDOI
HI-CFG: Construction by Binary Analysis and Application to Attack Polymorphism
TL;DR: This work introduces a new program representation, a hybrid information- and control-flow graph (HI-CFG), and gives algorithms to infer it from an instruction-level trace, and applies it to the task of generalizing an attack against a program whose inputs undergo complex transformations before reaching a vulnerability.
Posted Content
BACKDOORL: Backdoor Attack against Competitive Reinforcement Learning.
TL;DR: This paper demonstrates that an adversary agent can trigger the backdoor of the victim agent with its own action in two-player competitive RL systems and prototype and evaluate BACKDOORL in four competitive environments.
Book ChapterDOI
HookScout: proactive binary-centric hook detection
TL;DR: This paper implements two new attacks for Windows by exploiting two function pointers individually, proposes a new proactive hook detection technique, and develops a prototype, called HookScout, which can generate hook detection policy without access to the OS kernel source code.