Published papers such as 'Attention Sinks in Diffusion Language Models' (2025), 'Expected Attention: KV Cache Compression by Estimating Attention from Future Queries' (2025), 'Adaptive Computation Modules: Granular Conditional Computation For Efficient Inference' (AAAI 2025), etc.
Research Experience
Applied Agent Researcher at NVIDIA in Munich, Germany; Visiting researcher at Edinburgh NLP with Pasquale Minervini.
Education
PhD student in Data Science at Sapienza University of Rome, advisor Simone Scardapane; Erasmus student at Universidad Politecnica de Valencia, Spain.
Background
Research interests include efficient training and inference, AI interpretability, and adaptive & conditional computation methods. Specializes in both Computer Vision and Natural Language Processing.
Miscellany
Training to be a certified Life & Business Coach through the International Coaching Federation; Enjoys languages (even dead ones!) and teaches Ancient Greek and Latin to high school and college students.