Joël Mathys

Distributed Computing Group
Computer Engineering and Networks Laboratory (TIK)
Department Electrical Engineering (D-ITET)
Office ETZ G 63
ETH Zurich

Gloriastrasse 35
8092 Zurich
Switzerland

phone	+41 44 63 20417
fax	+41 44 63 21036
email

Publications

Long-Range Ising Model: A Benchmark for Long-Range Capabilities in Graph Learning
Joël Mathys, Henrik Christiansen, Federico Errica and Francesco Alesiani.
22nd International Workshop on Mining and Learning with Graphs (MLG@ECMLPKDD), Porto, Portugal, September 2025.
BibTeX
Learn to Jump: Adaptive Random Walks for Long-Range Propagation through Graph Hierarchies (oral)
Joël Mathys and Federico Errica.
ComBayNS Workshop @ IJCNN 2025 , Rome, Italy, June 2025.
BibTeX
Reinforcement Learning for Locally Checkable Labeling Problems
Florian Grötschla, Joël Mathys, Loïc Holbein and Roger Wattenhofer.
The Seventeenth Workshop on Adaptive and Learning Agents at AAMAS, Detroit, USA, May 2025.
BibTeX
Synthetic Data for Blood Vessel Network Extraction
Joël Mathys, Andreas Plesner, Jorel Elmiger and Roger Wattenhofer.
Will Synthetic Data Finally Solve the Data Access Problem? (SynthData@ICLR), Singapore, April 2025.
BibTeX
Beyond Interpolation: Extrapolative Reasoning with Reinforcement Learning and Graph Neural Networks
Niccolò Grillo, Andrea Toccaceli, Benjamin Estermann, Joël Mathys, Stefania Fresca and Roger Wattenhofer.
1st Workshop on Neural Reasoning and Mathematical Discovery (NEURMAD@AAAI25), Philadelphia, USA, March 2025.
BibTeX
GraphFSA: A Finite State Automaton Framework for Algorithmic Learning on Graphs
Florian Grötschla, Joël Mathys, Christoffer Raun and Roger Wattenhofer.
27th European Conference on Artificial Intelligence (ECAI), Santiago de Compostela, Spain, October 2024.
External BibTeX
CoRe-GD: A Hierarchical Framework for Scalable Graph Visualization with GNNs (Best poster award AT LOGML 2024)
Florian Grötschla, Joël Mathys, Robert Veres and Roger Wattenhofer.
12th International Conference on Learning Representations (ICLR), Vienna, Austria, May 2024.
External BibTeX
SALSA-CLRS: A Sparse and Scalable Benchmark for Algorithmic Reasoning
Julian Minder, Florian Grötschla, Joël Mathys and Roger Wattenhofer.
2nd Learning on Graphs Conference (LoG), Virtual, November 2023.
External BibTeX
SURF: A Generalization Benchmark for GNNs Predicting Fluid Dynamics
Stefan Künzli, Florian Grötschla, Joël Mathys and Roger Wattenhofer.
2nd Learning on Graphs Conference (LoG), Virtual, November 2023.
External BibTeX
Abstract Visual Reasoning Enabled by Language
Giacomo Camposampiero, Loïc Houmard, Benjamin Estermann, Joël Mathys and Roger Wattenhofer.
Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, June 2023.
External BibTeX
Learning Graph Algorithms With Recurrent Graph Neural Networks
Florian Grötschla, Joël Mathys and Roger Wattenhofer.
Workshop on Graphs and more Complex structures for Learning and Reasoning (GCLR@AAAI), Washington D.C., USA, February 2023.
External BibTeX
Hierarchical Graph Structures for Congestion and ETA Prediction
Florian Grötschla and Joël Mathys.
Traffic4cast@NeurIPS22, December 2022.
External Slides (PDF) Slides (PPT) BibTeX
Decentralized Graph Processing for Reachability Queries
Joël Mathys, Robin Fritsch and Roger Wattenhofer.
18th International Conference on Advanced Data Mining and Applications (ADMA), Brisbane, Australia, November 2022.
External Slides (PDF) Slides (PPT) BibTeX

Theses & Labs

	Title	Type	Contact/Supervisor	Assigned
Available Theses & Labs
	Data-Centric Generalization in Neural Algorithmic Reasoning	M/S	Joël Mathys, Florian Grötschla	available
Current Theses & Labs
	Title	Type	Contact/Supervisor	Assigned	Student(s)
	Advancing Neural Reasoning	S	Joël Mathys	HS 25
	Automated Foosball Commentary	B	Till Aczel, Joël Mathys	HS 25
	Detecting Blood Vessel Networks from 3D Scans	S	Andreas Plesner, Joël Mathys	HS 25
	External Graph Thesis	S	Joël Mathys	HS 25
	Linear Depth Graph Models	S	Joël Mathys, Saku Peltonen	HS 25
	Neural Blood Vessel Detection	S	Joël Mathys, Andreas Plesner	HS 25
	Pretraining GNNs	S	Saku Peltonen, Joël Mathys	HS 25
	SAT Solving with Graph Neural Networks	S	Saku Peltonen, Joël Mathys	HS 25
	Unleashing Forward Thinking in Language Models	M	Frédéric Berdoz, Joël Mathys	HS 25
	Mesh Optimization and Neural Closure Modeling Using Graph Neural Networks	M	Florian Grötschla, Joël Mathys	FS 25
Past Theses & Labs
	Title	Type	Contact/Supervisor	Assigned	Student(s)
	Automated Visual Foosball Tracking	B	Joël Mathys, Till Aczel	HS 24
	Expressive GNNs for SAT Solving through Substructure Counting	S	Saku Peltonen, Joël Mathys	HS 24
	Physics-Informed Graph Neural Networks for Airflow Simulations [confidential]	S	Joël Mathys, Florian Grötschla	HS 24
	FinGNN: Financial Graph Neural Networks [confidential]	M	Joël Mathys, Rudy Delacour	HS 24
	Wave Function Collapse for Graph Generation [confidential]	S	Till Aczel, Joël Mathys	HS 24
	Improving Investment Strategies with GNNs [confidential]	S	Florian Grötschla, Joël Mathys	FS 24
	Automated Vascular Graph Extraction [confidential]	M	Joël Mathys, Andreas Plesner	FS 24
	Reinforcement Learning for State-Based Algorithms on Graphs [confidential]	M	Florian Grötschla, Joël Mathys	FS 24
	Beyond Interpolation: Extrapolative Reasoning with Reinforcement Learning and Graph Neural Networks	G	Joël Mathys, Benjamin Estermann	FS 24
	Analyzing Deep Learning Architectures on Stock Prediction [confidential]	G	Florian Grötschla, Joël Mathys	FS 24	,
	Object-Centric Reasoning with Language and Code [confidential]	G	Benjamin Estermann, Joël Mathys	HS 23
	Exploring Activation Ensembles for Feed-Forward Neural Networks	B	Florian Grötschla, Joël Mathys	HS 23
	GNNs for TPU Graphs	S	Joël Mathys, Florian Grötschla	HS 23
	Train Small - Solve Big: Reinforcement Learning on Logic Puzzles using Graph Neural Networks [confidential]	M	Benjamin Estermann, Joël Mathys	HS 23
	Enhancing GNNs: An Exploration of Iterative Solving and Augmentation Techniques	B	Joël Mathys, Florian Grötschla	FS 23
	Enhancing the Building Blocks of Language-Based Visual Reasoning for ARC	M	Joël Mathys, Benjamin Estermann	FS 23
	Asynchronous GNNs [confidential]	G	Joël Mathys, Florian Grötschla	FS 23
	Finite State Algorithm Learning on Graphs [confidential]	M	Joël Mathys, Florian Grötschla	FS 23
	SALSA-CLRS: A Sparse and Scalable Benchmark for Algorithmic Reasoning [confidential]	S	Florian Grötschla, Joël Mathys	FS 23
	Reinforcement Learning of TSP Heuristics with Message Passing Neural Networks	G	Flint Xiaofeng Fan, Joël Mathys	FS 23	,
	A Generalisation Benchmark for Machine Learning Methods Predicting Fluid Flows [confidential]	B	Florian Grötschla, Joël Mathys	FS 23
	A Neural-Guided Program Synthesis Approach to the Abstraction and Reasoning Challenge [confidential]	G	Joël Mathys, Benjamin Estermann	HS 22
	Abstraction and Reasoning Challenge - Reasoning with Language [confidential]	G	Joël Mathys, Benjamin Estermann	HS 22	,
	Efficient Graph Drawing with GNNs using Overlay Graphs [confidential]	B	Florian Grötschla, Joël Mathys	HS 22
	Exploring Graph Neural Networks and Hierarchical Structures for Traffic Forecasting	B	Joël Mathys, Florian Grötschla	HS 22
	Real Time Public Transport App	S	Joël Mathys	FS 22

Joël Mathys

Publications

Theses & Labs

Available Theses & Labs

Current Theses & Labs

Past Theses & Labs