Mark Žnidar
Google Scholar | LinkedIn | GitHub | email | blog | If— | Artefacts
I am a Master's student in Advanced Computer Science at the University of Oxford, co-advised by Prof. Michael Bronstein (Oxford) and Prof. Jure Leskovec (Stanford). I was a Visiting Researcher at Stanford University and continue to collaborate with the Leskovec group. I am an ASEF Junior Research Fellow. Previously, I completed a BSc in Mathematics and Computer Science at the University of Ljubljana and spent a fruitful year doing research in the industry at Teads, pushing the frontier of real-time bidding and large-scale recommender systems.
My research interests span the theoretical and applied foundations of machine learning, with a focus on developing principled and scalable algorithms for learning on structured data and graph neural networks. I am also exploring the design of foundation models for relational domains.
🏀 Interesting fact: I played basketball professionally and was the youngest player in the highest national league and in international competitions.
news
| Feb 2026 | Publication accepted at DATA-FM @ ICLR'26: RelBench v2: A Large-Scale Benchmark and Relational Data Repository. |
|---|---|
| Jan 2026 | Relational Transformer accepted to ICLR 2026 (oral). |
| Oct 2025 | SAP CTO highlights the importance of our Relational Transformer. |
| Oct 2025 | Preprint: Relational Transformer: Toward Zero-Shot Foundation Models for Relational Data. |
| Jun 2025 | Started as Visiting Researcher at Stanford University (Leskovec group). |
| Jan 2025 | Selected as ASEF Junior Research Fellow. |
| Jul 2024 | Started research on real-time bidding systems at Teads. |
pre-prints
-
Multi-optimizer Deep&Cross at Industrial Scale
Mark Žnidar, Assaf Klein, Blaž Škrlj
Under review, European Conference on Information Retrieval (ECIR) Industry Track, 2026
publications
-
RelBench v2: A Large-Scale Benchmark and Relational Data Repository
Justin Gu, Rishabh Ranjan, Charilaos I. Kanatsoulis, Haiming Tang, Martin Jurkovič, Valter Hudovernik, Mark Žnidar, Pranshu Chaturvedi, Parth Shroff, Fengyu Li, Jure Leskovec
DATA-FM @ ICLR, 2026
arxiv -
Relational Transformer: Toward Zero-Shot Foundation Models for Relational Data Also: Spotlight presentation at NeurIPS AI4TD workshop
Rishabh Ranjan, Valter Hudovernik, Mark Žnidar, Charilaos Kanatsoulis, Roshan Upendra, Mahmoud Mohammadi, Joe Meyer, Tom Palczewski, Carlos Guestrin, Jure Leskovec
International Conference on Learning Representations (ICLR), 2026
arxiv -
Multi-Modal Embedding Fusion for Scalable Context-First CTR Rising Star Award
Mark Žnidar, Blaž Mramor, Natalia Silberstein, Anže Alič, Martin Jakomin, Blaž Škrlj
2025 IEEE International Conference on Data Mining Workshops (ICDMW)
ICDMW proceedings -
Utility of embeddings in multi-modal models for click-through rate prediction
Mark Žnidar, Blaž Škrlj, Martin Jakomin, Marko Robnik-Šikonja
Bachelor's Thesis, University of Ljubljana, 2025
thesis