Our work on how LLMs store relations selected as NeurIPS Spotlight paper

Our paper The Structure of Relation Decoding Linear Operators in Large Language Models by Miranda Anna Christ, Adrián Csiszárik, Gergely Becsó, and Dániel Varga was accepted at the NeurIPS 2025 conference as a Spotlight paper (~3% of submissions).

Healthcare

Prevention & Prediction

Patient Pathways

Patient Pathway Mission

Leveraging Hungary’s healthcare data assets for prevention, prediction, and decision support.

Közelebb a matematikához - Miklós Dezső és Csiszárik Adrián - HUN-REN Rényi Alfréd Matematikai Kutatóintézet

Podcast
June 25, 2025
podcasts
Miként támogathatja az AI a hazai egészségügyi ellátórendszer hatékonyabb működését? Hogyan segíthet a Mesterséges Intelligencia az orvosnak a betegségek korai felismerésében, a kezelés megtervezésében, egyáltalán abban, hogy a betegről rendelkezésre álló óriási mennyiségű és minőségű adat, információ átlátható legyen? A műsor vendégei a HUN-REN Rényi Alfréd Matematikai Kutatóintézet igazgatóhelyettese, Miklós Dezső és a Rényi Intézet Mesterséges Intelligencia kutatócsoportjának tagja, az említett kutatás szakmai vezetője Csiszárik Adrián. A Közelebb a matematikához című műsor ezúttal bemutatja azt az évek óta folyó munkát, amely során a Rényi kutatói átfogó egészségügyi életútelemző adatplatformot építenek.
June 25, 2025

Diverse beam search to find densest-known planar unit distance graphs

Publication
June 13, 2025
publications
This paper addresses the problem of determining the maximum number of edges in a unit distance graph (UDG) of n vertices using computer search. An unsolved problem of Paul Erdős asks the maximum number of edges 𝑢⁡(𝑛) a UDG of n vertices can have. Those UDGs that attain 𝑢⁡(𝑛) are called “maximally dense.” In this paper, we seek to demonstrate a computer algorithm to generate dense UDGs for vertex counts up to at least 100.
June 13, 2025

Zsolt Zombori

Better Exploration for Symbolic Supervision

Blog Post
June 2, 2025
posts
Despite the tremendous success that deep learning has shown in the past decade, there are certain application domains for which deep learning has traditionally been considered as not suitable. In particular, deep learning is notoriously known to be very data inefficient and to provide no guarantees with respect to the behaviour of the trained system. As a result of these limitations, the usage of neural networks in small data and safety critical domains has so far remained rather restricted.
June 2, 2025

Piercing intersecting convex sets

Publication
April 1, 2025
publications
Assume two finite families A and B of convex sets in R3 have the property that A ∩ B ≠ ∅ for every A ∈ A and B∈B. Is there a constant γ > 0 (independent of A and B) such that there is a line intersecting γ|A| sets in A or γ|B| sets in B? This is an intriguing Helly-type question from a paper by Martínez, Roldan and Rubin. We confirm this in the special case when all sets in A lie in parallel planes and all sets in B lie in parallel planes; in fact, one of the two families has a transversal by a single line.
April 1, 2025
The Team
The AI group at the institute brings together experts with backgrounds in both industry and academia. We place equal emphasis on theoretical foundations, thorough experimentation, and practical applications. Our close collaboration ensures a continuous exchange of knowledge between scientific research and applied projects.
Balázs Szegedy
Mathematical Theory
Attila Börcs, PhD
NLP, Modeling, MLOps
Adrián Csiszárik
Representation Learning, Foundations
Győző Csóka
NLP, MLOps
Domonkos Czifra
NLP, Foundations
Botond Forrai
Modeling
Péter Kőrösi-Szabó
Modeling
Gábor Kovács
NLP, Modeling
Judit Laki, MD PhD
Healthcare
Márton Muntag
Time Series, NLP, Modeling
Dávid Terjék
Generalization, Mathematical Theory
Dániel Varga
Foundations, Computer aided proofs
Pál Zsámboki
Reinforcement Learning, Geometric Deep Learning
Zsolt Zombori
Formal Reasoning
Péter Ágoston
Combinatory, Geometry
Beatrix Mária Benkő
Representation Learning
Jakab Buda
NLP
Diego González Sánchez
Generalization, Mathematical Theory
Melinda F. Kiss
Representation Learning
Ákos Matszangosz
Topology, Foundations
Alex Olár
Foundations
Gergely Papp
Modeling
Open Positions
The Rényi AI group is actively recruiting both theorists and practitioners.
Announcement: July 1, 2025
Deadline: rolling
applications
Rényi Institute is seeking Machine Learning Engineers to join our AI Research & Development team. Preferred Qualifications: • MLOps experience (especially in cloud environments) • Industry experience working on ML solutions
Announcement: July 1, 2025
Deadline: rolling
theory, applications
Rényi Institute is seeking Research Scientists to join our AI Research & Development team. You will have the privilege to work at a renowned academic institute and do what you love: do research and publish in the field of machine learning / deep learning.
Rényi AI - Building bridges between mathematics and artificial intelligence.