Emile van Krieken

University of Edinburgh

Emile.van.Krieken@ed.ac.uk

I am a postdoc in the NLP group and the APRIL lab at the University of Edinburgh under the ELIAI program. I obtained a PhD with distinction (cum laude) in Artificial Intelligence at the Vrije Universiteit Amsterdam in 2024, where I am also a visiting researcher in the Learning and Reasoning group.

My research combines machine learning with symbolic reasoning, or “Neurosymbolic Learning”. My work focuses on the fundamental understanding of such combinations, including optimisation properties, characterisations of expressiveness, and scalability. Currently, I am particularly interested in methods for Neurosymbolic Learning in Generative AI models like LLMs and Diffusion Models. Another focus is the accessibility of Neurosymbolic Learning, to which end I lead the development of the ULLER Python library

I am also interested in Personal Knowledge Management and developed Juggl, a plugin for Obsidian.md that adds a customizable graph view. Other plugins include Supercharged Links. I also composed some music: You can listen to some songs here.

I am open for academic opportunities in Machine Learning and AI, preferably in or close to the Netherlands.

news

Nov 21, 2024	I will be Program Chair of the NeSy 2025 conference in Santa Cruz, USA from 8-10 September 2025.
Sep 09, 2024	Our paper on ULLER, a general-purpose NeSy language, is accepted as a spotlight oral presentation at NeSy 2024. In addition, I’m presenting my ICML paper as a spotlight oral, and we are giving a tutorial on ULLER on Thursday the 12th of September.
Jun 27, 2024	I am giving a keynote talk at the Differentiable Almost Everything ICML 2024 workshop.
May 02, 2024	Our paper “On the Independence Assumption in Probabilistic Neurosymbolic Learning” is accepted at ICML 2024.
Apr 26, 2024	Our paper “A Neuro-Symbolic Benchmark Suite for Concept Quality and Reasoning Shortcuts ” is accepted at the NeurIPS 2025 Datasets and Benchmarks track.

selected publications

NeSy (SPOTLIGHT)

ULLER: A Unified Language for Learning and Reasoning

Emile van Krieken, Samy Badreddine, Robin Manhaeve, and 1 more author

In 18th International Conference on Neural-Symbolic Learning and Reasoning, 2024

Abs arXiv

The field of neuro-symbolic artificial intelligence (NeSy), which combines learning and reasoning, has recently experienced significant growth. There now are a wide variety of NeSy frameworks , each with its own specific language for expressing background knowledge and how to relate it to neural networks. This heterogeneity hinders accessibility for newcomers and makes comparing different NeSy frameworks challenging. We propose a unified language for NeSy, which we call ULLER, a Unified Language for LEarning and Reasoning. ULLER encompasses a wide variety of settings, while ensuring that knowledge described in it can be used in existing NeSy systems. ULLER has a neuro-symbolic first-order syntax for which we provide example semantics including classical, fuzzy, and probabilistic logics. We believe ULLER is a first step towards making NeSy research more accessible and comparable, paving the way for libraries that streamline training and evaluation across a multitude of semantics, knowledge bases, and NeSy systems.
NeurIPS

A-NeSI: A Scalable Approximate Method for Probabilistic Neurosymbolic Inference

Emile van Krieken, Thiviyan Thanapalasingam, Jakub Tomczak, and 2 more authors

In Advances in Neural Information Processing Systems, 2023

Abs arXiv Code

We study the problem of combining neural networks with symbolic reasoning. Recently introduced frameworks for Probabilistic Neurosymbolic Learning (PNL), such as DeepProbLog, perform exponential-time exact inference, limiting the scalability of PNL solutions. We introduce Approximate Neurosymbolic Inference (A-NeSI): a new framework for PNL that uses neural networks for scalable approximate inference. A-NeSI 1) performs approximate inference in polynomial time without changing the semantics of probabilistic logics; 2) is trained using data generated by the background knowledge; 3) can generate symbolic explanations of predictions; and 4) can guarantee the satisfaction of logical constraints at test time, which is vital in safety-critical applications. Our experiments show that A-NeSI is the first end-to-end method to solve three neurosymbolic tasks with exponential combinatorial scaling. Finally, our experiments show that A-NeSI achieves explainability and safety without a penalty in performance.
ICML

On the Independence Assumption in Neurosymbolic Learning

Emile van Krieken, Pasquale Minervini, Edoardo M Ponti, and 1 more author

In , 2024
NeurIPS

Storchastic: A Framework for General Stochastic Automatic Differentiation

Emile van Krieken, Jakub Tomczak, and Annette Ten Teije

In Advances in Neural Information Processing Systems, 2021