Adversarial robustness
- ZOO: black-box attacks on deep neural networks (ACM AISec 2017)
- Co-author of the book "Adversarial Robustness for Machine Learning"
Research network
A network of researchers.
Researchers from around the world work with our crew, alongside our partner laboratory. They give talks, share their work and open the door to real research.
Research partner
L'Empreinte
ESTIN's student journal · Algeria
ESTIN's student journal on technology, ideas and the people who build them. INSIGHT members publish their research papers in its pages.
Research partner
Research Laboratory Dz
R&D Project · Algeria
Empowers academics, professionals and companies through research, innovation and scientific projects.
Researchers
7 researchers · 4 countries
-
Welcome Day guest -
ESTIN, Heriot-Watt University · Algeria
Prof. Hani Ragab Hassen
Professor, Information Security
Profile ↗ -
North Carolina A&T State University · USA
Dr. Islam Obaidat
Assistant Professor, Computer Systems Technology
Profile ↗ -
North Carolina A&T State University · USA
Prof. Muhammad Khan
Assistant Professor, Built Environment
Profile ↗ -
University of Waterloo · Canada
Dr. Mohamed Camil Belhadjoudja
Postdoctoral Fellow, Control Theory
Profile ↗ -
-
James Madison University · USA
Ford Prior
Senior Staff Engineer and Tech Evangelist, GEICO
Profile ↗ - Get in touch →
Are you a researcher?
Mentor a project team, give a talk, or co-supervise a student project.
Log 02 · Featured researcher
Dr. Pin-Yu Chen
Principal Research Scientist · IBM Research
Principal Research Scientist at the IBM Thomas J. Watson Research Center (Yorktown Heights, New York). His long-term goal is to build trustworthy machine learning systems, with a focus on AI safety and robustness.
- Chief Scientist, RPI–IBM AI Research Collaboration
- PI of MIT–IBM Watson AI Lab projects
- PhD in Electrical Engineering & Computer Science and MA in Statistics, University of Michigan (2016)
- IEEE Fellow · IJCAI Computers & Thought Award 2023 · IBM Master Inventor
Building trustworthy AI
Safety of large language models
- Gradient Cuff: detecting jailbreak attempts on LLMs
- Attention Tracker: detecting prompt-injection attacks (NAACL Findings 2025)
Looking inside models
- vLLM Hook (2026): open-source plug-in to inspect and steer the internal states of LLMs served on vLLM
- Contributor to IBM's AI Explainability 360 toolkits
Open science
- Contributor to IBM's Adversarial Robustness Toolbox (ART)
- Co-author of the book "Introduction to Foundation Models"
Why his work matters to our crew
- AI now runs in critical systems. It must hold up under attack.
- AI agents read untrusted content. Prompt injection is a real threat.
- Looking inside models gives defenders new tools.
Log 03 · Guest talks
Talks with our crew.
Event · 4 NOV 2026
Welcome Day 2026: INSIGHT sets sail
The official start of INSIGHT, with our four research poles and a live talk by Dr. Pin-Yu Chen from IBM Research.
Read the recap →