Michael M. Bronstein

LLMs Can Hide Text in Other Text of the Same Length

Efficient Learning on Large Graphs using a Densifying Regularity Lemma

Scaling Atomistic Protein Binder Design with Generative Pretraining and Test-Time Compute

Efficient Regression-based Training of Normalizing Flows for Boltzmann Generators

Carré du champ flow matching: better quality-generalisation tradeoff in generative models

OXtal: An All-Atom Diffusion Model for Organic Crystal Structure Prediction

Towards Quantifying Long-Range Interactions in Graph Machine Learning: a Large Graph Dataset and a Measurement

Planner Aware Path Learning in Diffusion Language Models Training

Attention Sinks and Compression Valleys in LLMs are Two Sides of the Same Coin

Flock: A Knowledge Graph Foundation Model via Learning on Random Walks

TGM: A Modular and Efficient Library for Machine Learning on Temporal Graphs

ResCP: Reservoir Conformal Prediction for Time Series Forecasting

HYPER: A Foundation Model for Inductive Link Prediction with Knowledge Hypergraphs

Generalised Flow Maps for Few-Step Generative Modelling on Riemannian Manifolds

Self-Exploring Language Models for Explainable Link Forecasting on Temporal Graphs via Reinforcement Learning

gLSTM: Mitigating Over-Squashing by Increasing Storage Capacity

Learning Escorted Protocols For Multistate Free-Energy Estimation

MarS-FM: Generative Modeling of Molecular Dynamics via Markov State Models

Learning Inter-Atomic Potentials without Explicit Equivariance

GraphBench: Next-generation graph learning benchmarking

Bringing Graphs to the Table: Zero-Shot Node Classification via Tabular Foundation Models

Are Large Language Models Good Temporal Graph Learners?

Topological Blindspots: Understanding and Extending Topological Deep Learning Through the Lens of Expressivity

Progressive Inference-Time Annealing of Diffusion Models for Sampling from Boltzmann Densities

Drag-and-Drop LLMs: Zero-Shot Prompt-to-Weights

Bundle Neural Network for message diffusion on graphs

Equivariance Everywhere All At Once: A Recipe for Graph Foundation Models

Over-squashing in Spatiotemporal Graph Neural Networks

Curly Flow Matching for Learning Non-gradient Field Dynamics

On Vanishing Gradients, Over-Smoothing, and Over-Squashing in GNNs: Bridging Recurrent and Graph Learning

On Measuring Long-Range Interactions in Graph Neural Networks

Fully-inductive Node Classification on Arbitrary Graphs

GradMetaNet: An Equivariant Architecture for Learning on Gradients

Gradient Variance Reveals Failure Modes in Flow-Based Generative Models

Amortized Sampling with Transferable Normalizing Flows

Homomorphism Counts as Structural Encodings for Graph Learning

Neural Spacetimes for DAG Representation Learning

Why do LLMs attend to the first token?

Multi-domain Distribution Learning for De Novo Drug Design

Balancing Efficiency and Expressiveness: Subgraph GNNs with Walk-Based Centrality

Scalable Equilibrium Sampling with Sequential Boltzmann Generators

Link Prediction with Relational Hypergraphs

Steering Masked Discrete Diffusion Models via Discrete Denoising Posterior Prediction

Balancing Efficiency and Expressiveness: Subgraph GNNs with Walk-Based Centrality

Understanding Virtual Nodes: Oversquashing and Node Heterogeneity

Supercharging Graph Transformers with Advective Diffusion

Learning Symmetries through Loss Landscape

Laplace-Transform-Filters render spectral Graph Neural Networks transferable

Scalable Message Passing Neural Networks: No Need for Attention in Large Graph Representation Learning

Wavelet-based Graph Convolution via Chebyshev Decomposition

A General Graph Spectral Wavelet Convolution via Chebyshev Order Decomposition

How Expressive are Knowledge Graph Foundation Models?

DyGMamba: Efficiently Modeling Long-Term Temporal Dependency on Continuous-Time Dynamic Graphs with State Space Models

Towards Generalization under Topological Shifts: A Diffusion PDE Perspective