Home
My name is Eleonora Giunchiglia and I am an Assistant Professor at Imperial College London affiliated with Imperial-X and the Electrical and Electronic Engineering Department. Previously, I completed my Post-doc at TU Wien, and I obtained my PhD at the University of Oxford in 2022. Regarding my education, I have a Bachelor in Computer Engineering and a MSc in Data Science and Engineering from the University of Genova and a MSc in Computer Science from the University of Oxford.
I’m passionate about building machine learning models that are not just powerful, but safe by-design: guaranteed to satisfy formally expressed requirements. I strongly believe that requirements encode valuable background knowledge about a problem, and that models should not only respect them, but actively exploit them to learn more efficiently from fewer datapoints and parameters. To this end, I develop novel techniques for embedding formal constraints directly into the architecture and training process of neural networks. So far, I’ve explored exciting applications in hierarchical classification, autonomous driving, tabular data generation, and natural language generation—with many more to come!
News
- November 2025 I will be giving one of the Frontiers in AI talks at ECAI 2025
- July 2025 I will be giving an invited talk at the 8th International Symposium on AI Verification in Zagreb (Croatia)
- July 2025 I will be giving a lecture at the Imperial Policy Forum
- June 2025 Our paper TRIDENT: Temporally Restricted Inference via DFA-Enhanced Neural Traversal is now on arxiv
- March 2025 Our paper A survey on tabular data generation: Utility, alignment, fidelity, privacy, and beyond is now on arxiv
- February 2025 Our paper Beyond the convexity assumption: Realistic tabular data generation under quantifier-free real linear constraints has been accepted at ICLR 2025
- February 2025 Together with Emile we have given a lecture on Learning with Constraints at Dagstuhl