EMNLP 2026 · Findings 01
From Layers to Submodules: Rethinking Granularity in Replacement-Based LLM Compression
Submodule-level compression with lightweight fitted residual bypasses.
AI Researcher
Applied AI for efficient LLMs.
Hi, I’m Elia — an Applied AI Researcher and fourth-year PhD candidate at the University of Trento, with experience across academia and industry.
My research focuses on model compression, pruning, sparsity, data curation, and post-training for large language models.
01 Model compression
02 Efficient LLMs
03 Post-training
Presented at EMNLP 2026
Latest work
Now · Turin, Italy
Post-training at Alexa AI.
I am currently an Applied Scientist Intern at Amazon in Turin, working with Alexa AI on post-training.
Apr–Jun 2026 · Seville, Spain
From papers to biomedical discovery.
Between April and June 2026, I joined Bristol Myers Squibb (CITRE) in Seville as a Visiting Scientist. I worked with Dr. Santiago Egea Gomez and Dr. Hannah Pliner on generating gene regulatory networks from scientific literature using large language models.
Since 2022 · Trento, Italy
PhD student at the University of Trento.
I have been a PhD student in Computer Science at the University of Trento since 2022, researching efficient deep learning, model compression, pruning, and sparsity.