PI of DUCK Lab
We build AI systems that don't just learn from data, but reason over it. Systems that acknowledge what they don't know, respect the rules of the world, and draw on structured knowledge to make better decisions.
Four words sit at the heart of everything we do. Data. Uncertainty. Constraints. Knowledge. At the DUCK Lab we build AI systems that don't compromise between being powerful and being safe — a small, tight-knit team at Imperial College London united by the belief that rigor and creativity belong in the same sentence.
Eleonora will give an invited talk at CILC 2026, the 41st Italian Conference on Computational Logic.
🚀 We are thrilled to announce the release of Pythagoras-Prover — a compute-efficient family of Lean theorem provers! Strong results at just 4B parameters, powered by Augmented Lean Formalisation. The models are now available on Hugging Face 🤗
On 9 May, Eleonora spoke at the OxBridge Women in CS Conference — a particularly special one for her, as she was herself an OxWoCS member during her PhD!
Our paper Can I Have Your Order? Monte-Carlo Tree Search for Slot Filling Ordering in Diffusion Language Models has been accepted at ICML 2026.
Our paper PiCSAR: Probabilistic Confidence Selection And Ranking for Reasoning Chains has been accepted at ACL (Findings) 2026.