Search Results for author: Karolis Martinkus

Found 9 papers, 5 papers with code

On Isotropy Calibration of Transformer Models

no code implementations insights (ACL) 2022 Yue Ding, Karolis Martinkus, Damian Pascual, Simon Clematide, Roger Wattenhofer

Different studies of the embedding space of transformer models suggest that the distribution of contextual representations is highly anisotropic - the embeddings are distributed in a narrow cone.

Automating Rigid Origami Design

no code implementations20 Nov 2022 Jeremia Geiger, Karolis Martinkus, Oliver Richter, Roger Wattenhofer

While rigid origami has shown potential in a large diversity of engineering applications, current rigid origami crease pattern designs mostly rely on known tessellations.

Diffusion Models for Graphs Benefit From Discrete State Spaces

1 code implementation4 Oct 2022 Kilian Konstantin Haefeli, Karolis Martinkus, Nathanaël Perraudin, Roger Wattenhofer

Denoising diffusion probabilistic models and score matching models have proven to be very powerful for generative tasks.

Denoising

Agent-based Graph Neural Networks

1 code implementation22 Jun 2022 Karolis Martinkus, Pál András Papp, Benedikt Schesch, Roger Wattenhofer

AgentNet is inspired by sublinear algorithms, featuring a computational complexity that is independent of the graph size.

Graph Classification

SPECTRE: Spectral Conditioning Helps to Overcome the Expressivity Limits of One-shot Graph Generators

1 code implementation4 Apr 2022 Karolis Martinkus, Andreas Loukas, Nathanaël Perraudin, Roger Wattenhofer

We approach the graph generation problem from a spectral perspective by first generating the dominant parts of the graph Laplacian spectrum and then building a graph matching these eigenvalues and eigenvectors.

Graph Generation Graph Matching

DropGNN: Random Dropouts Increase the Expressiveness of Graph Neural Networks

1 code implementation NeurIPS 2021 Pál András Papp, Karolis Martinkus, Lukas Faber, Roger Wattenhofer

In DropGNNs, we execute multiple runs of a GNN on the input graph, with some of the nodes randomly and independently dropped in each of these runs.

Graph Classification Graph Regression

On Isotropy Calibration of Transformers

no code implementations27 Sep 2021 Yue Ding, Karolis Martinkus, Damian Pascual, Simon Clematide, Roger Wattenhofer

Different studies of the embedding space of transformer models suggest that the distribution of contextual representations is highly anisotropic - the embeddings are distributed in a narrow cone.

Scalable Graph Networks for Particle Simulations

1 code implementation14 Oct 2020 Karolis Martinkus, Aurelien Lucchi, Nathanaël Perraudin

However, the dynamics of many real-world systems are challenging to learn due to the presence of nonlinear potentials and a number of interactions that scales quadratically with the number of particles $N$, as in the case of the N-body problem.

Cannot find the paper you are looking for? You can Submit a new open access paper.