Limin Yang, CS PhD Student at UIUC

Limin Yang (杨利民)

I am a fifth-year Ph.D. candidate of Computer Science at UIUC.
I focus on machine learning security and explainable AI.
My advisor is Prof. Gang Wang.

I'm on the industry job market! Please email me if you think I'd be a good fit for your team.

Contact: liminy2@illinois.edu
Office: Room 4111, Thomas M. Siebel Center, Urbana, IL 61801.

News

03/10/2023	Our Jigsaw Puzzle backdoor attack paper got accepted in IEEE S&P'23 (Oakland). Another paper also got conditionally accepted. Big thanks to everyone and see you at San Francisco!
02/21/2023	Just passed my preliminary exam. Thanks all my committee members!
08/19/2022	Submitted a paper to IEEE S&P (Oakland) 2023! Thanks a lot to my advisor and collaborators!
05/23/2022	First day at IBM Research (New York) as a visiting scholar. I work with Muhammed Fatih Bulut and Constantin Adam on network intrusion detection.
05/17/2021	First day at TikTok (ByteDance) as a Security Engineering Intern (focus: spam detection). Thanks for the opportunity, Dr. Yu Ding and Dr. Hui Cao!
04/09/2021	Just passed my qualifier exam. Now I'm officially a Ph.D. candidate!
02/24/2021	Our paper got accepted to the DLS workshop. We release a new PE malware dataset for ML research [download link].
11/30/2020	I will be serving as a Shadow PC at Oakland 2021.
09/30/2020	Our USENIX Security'21 paper got accepted! Thanks for everyone who helped me!

Publications

Conference Papers

[IEEE S&P'23]	Jigsaw Puzzle: Selective Backdoor Attack to Subvert Malware Classifier Limin Yang, Zhi Chen, Jacopo Cortellazzi, Feargus Pendlebury, Kevin Tu, Fabio Pierazzi, Lorenzo Cavallaro, Gang Wang TL;DR: Jigsaw Puzzle backdoor attack can selectively protect one specific malware family but not other malware, leaving smaller footprints and thus bypassing many existing backdoor defense methods.
[USENIX Security'21]	CADE: Detecting and Explaining Concept Drift Samples for Security Applications Limin Yang, Wenbo Guo, Qingying Hao, Arridhana Ciptadi, Ali Ahmadzadeh, Xinyu Xing, Gang Wang PDF / Supplementary PDF / Code / Slides / Video / Artifact Evaluated TL;DR: Contrastive learning can help detect drifting samples from previously unseen classes and distance-based explanation can find out why such drifting samples are different from existing classes.
[USENIX Security'20]	Measuring and Modeling the Label Dynamics of Online Anti-Malware Engines Shuofei Zhu, Jianjun Shi, Limin Yang, Boqin Qin, Ziyi Zhang, Linhai Song, Gang Wang PDF / Data / Artifact Evaluated TL;DR: How to aggregate VirusTotal labels more properly?
[IMC'19]	Opening the Blackbox of VirusTotal: Analyzing Online Phishing Scan Engines Peng Peng, Limin Yang, Linhai Song, Gang Wang PDF / Data TL;DR: Control simple phishing sites to see how bad and inconsistent of the URL detection provided by VirusTotal and security vendors' own APIs.
[USENIX Security'18]	Understanding the Reproducibility of Crowd-reported Security Vulnerabilities Dongliang Mu, Alejandro Cuevas, Limin Yang, Hang Hu, Xinyu Xing, Bing Mao, Gang Wang PDF / Data TL;DR: Provide quantitive evidence on the prevalence of missing information in vulnerability reports and low reproducibility by manually reproduce 368 memory corruption bugs.
[GLOBECOM'17]	VulDigger: A Just-in-Time and Cost-Aware Tool for Digging Vulnerability-Contributing Changes Limin Yang, et al. PDF TL;DR: Use metadata and heuristics to predict whether a git commit introduced a vulnerability.

Journal / Workshop Papers

[DLS'21]	BODMAS: An Open Dataset for Learning based Temporal Analysis of PE Malware Limin Yang, Arridhana Ciptadi, Ihar Laziuk, Ali Ahmadzadeh, Gang Wang PDF / Slides / Data / Code TL;DR: We collaborate with Blue Hexagon to release an open Windows PE malware dataset with timestamped metadata as well as curated malware family labels.
[SafeThings'20]	A Case Study of the Security Vetting Process of Smart-home Assistant Applications Hang Hu, Limin Yang, Shihan Lin, Gang Wang PDF TL;DR: Verify that replay attack and SQL injection are feasible in Amazon Alexa / Google Home apps.
[PPNA'17]	Characterizing User Behaviors in Location-based Find-and-Flirt Services: Anonymity and Demographics Minhui Xue, Limin Yang, Keith W. Ross, Haifeng Qian PDF TL;DR: Who used the WeChat "People Nearby" feature more often?

Internships

IBM Research	Visiting Scholar (Research Intern) , New York, US, 05/2022 – 08/2022 Mentors: Dr. Muhammed Fatih Bulut and Dr. Constantin Adam
TikTok	Security Engineering Intern, California, US, 05/2021 – 08/2021 Mentors: Dr. Dazhuo Li, Dr. Hamed Ashouri, and Dr. Hui Cao
Penn State	Research Intern, Pennsylvania, US, 09/2017 – 02/2018 Mentors: Dr. Xinyu Xing and Dr. Gang Wang
Peking University	Exploit Intern, Beijing, China, 07/2015 – 08/2015 Mentors: Dr. Yu Ding and Dr. Xinhui Han.

Awards

2021	CCS Student Conference Grant
2017	ECNU Graduate Student Overseas Research Scholarship
2013-2015	ECNU Top-notch Innovative Personnel Training Plan

Teaching

2022 Fall	CS-463 Computer Security II, UIUC, Teaching Assistant
2019 Spring	CS-4264 Principles of Computer Security, Virginia Tech, Teaching Assistant
2018 Fall	CS-3114 Data Structures and Algorithms, Virginia Tech, Teaching Assistant

Professional Services

[Oakland'21]	IEEE Symposium on Security and Privacy, Student PC
[Patterns'21]	Patterns (a data science journal by Cell Press), Reviewer

Upcoming Deadlines

Oakland'24	April 13, August 3, December 6, 2023. San Francisco, CA.
USENIX Sec'24	June 6, October 17, 2023; February 8, 2024. Philadelphia, PA.
CCS'23	~~January 19,~~ May 4, 2023. Copenhagen, Denmark.
NDSS'24	April 19, June 28, 2023. San Diego, CA.

Miscellaneous

I'm also a minimalism photographer: whyisyoung.com

Feel free to steal this website's source code, just add a link back to my website and credit Jon Barron and me. Please remove any code that contains "UA-45384067-2" and "GTM-N2T82H2" as they are my Google Analytic tracking ID and you do not want on your own website.