The 2nd AEGIS Symposium on Cyber Security (2023)

Preliminary Program

The preliminary program is now available. The schedule and topic may be subject to change upon speakers' request and availability, but we aim at settling the schedule of talks 1 week prior to the planned date.

The Zoom link will be available 1 week prior to the talk. Please directly click on the talk listed below to enter the conference zoom.

Upcoming (click on the card for entering Zoom):

Scheduled:

Post Events (click on the card for video if publicized):

Session 1: ML-LLM [ENGLISH]

Understanding the Capability for LLM to Reason Program Behavior [Video]

Kexin Pei, Columbia University
Session 2: Cyber Infrastructure [ENGLISH]

Secure and Privacy-preserving AI Foundation Models [Video]

Hongbin Liu, Duke University
Session 2: Cyber Infrastructure [ENGLISH]

Designing Decentralized Blockchain Mechanisms

Ke Wu, Carnegie Mellon University
Session 1: ML-LLM

Measuring the Reliability of ChatGPT [ENGLISH]

Xinyue Shen, CISPA Helmholtz Center for Information Security [Video]
Session 2: Cyber Infrastructure

Operationalizing Machine Learning for Networks [Video]

Shinan Liu, University of Chicago
Session 1: ML-LLM [ENGLISH]

Certifiably Robust Defense against Adversarial Patch Attacks [Video]

Chong Xiang, Princeton University
Session 1: ML-LLM

Certified Trustworthiness in ML

Linyi Li, University of Illinois Urbana-Champaign [Video]
Session 3: Novel Attacks [ENGLISH]

Modeling Performance Issues for Shielding the Cloud and Web Applications

Yinxi Liu, Chinese University of Hong Kong
Session 3: Novel Attacks

On the Pitfalls and Emerging Threats in Super App Paradigm

Yuqing Yang, The Ohio State University
Session 4: Cyber Threats

Vulnerabilities in Domain Name Delegation and Revocation [Video]

Xiang Li, Tsinghua University
Session 4: Cyber Threats

Instant Synthesis of Smart Contract Counterattacks [Video]

Zhuo Zhang, Purdue University
Session 4: Cyber Threats

Learned Graph Databases for Provenance Tasks [Video]

Hailun Ding, Rutgers University