I'm Will Kalikman, a Master's student in computer science at ETH Zurich. I'm currently doing a thesis on gradient leakage attacks with the SRI lab and INSAIT in the secure and trustworthy AI group under the supervision of Martin Vechev. I'm fortunate to be advised by Dimitar Dimitrov and Ivo Petrov.
Before starting my thesis, I spent most of 2025 researching optimization-based adversarial attacks on language models. My research interests are in AI safety and what gradient-based optimization over inputs can reveal about trained models.
I received my undergraduate degree from Dartmouth College (BS CS & BA Philosophy).
All pictures on the site taken by yours truly.
You can reach me at: w[lastname]atgmail dot com.
-
A paper on augmenting source text to make machine translation harder.
-
My first research project in adversarial ML.
-
Why I made this site.
-
On building interpretable ML tools at Coinbase.
-
Old notes on what we actually want when we ask for explanations of ML systems.