Hi, I'm Baimam Boukar 👋

I am a research and software engineer working on mechanistic interpretability and the science of evaluations for advanced language models. At Jinesis AI Lab, University of Toronto, I study when probes, sparse autoencoders, and causal interventions generalize across model families, layers, tasks, and data distributions. I earned my MSc in Information Technology, specializing in Applied Machine Learning, from Carnegie Mellon University Africa.