Frederik Benzing

Research Interests

I am a PhD student in Prof. Dr. Angelika Steger's neuroscience group. Before starting my PhD in Computer Science, I studied Mathematics at the University of Cambridge and at ETH Zurich. I am interested in how research in machine learning and neuroscience can benefit from each other.

Publications and Preprints

Gradient Descent on Neurons and its Link to Approximate Second-order Optimisation   [code]   [5min Video]
F. Benzing
ICML, 2022; Previous version was presented as a spotlight at the OPT-ML Workshop at NeurIPS 2021

Unifying Regularisation Methods for Continual Learning   [code]   [12min Video]  
F. Benzing
AISTATS 2022, Oral Presentation

Presynaptic stochasticity improves energy efficiency and helps alleviate the stability-plasticity dilemma
S. Schug*, F. Benzing*, A. Steger
eLife, 10/2021

Optimal Kronecker-Sum Approximation of Real Time Recurrent Learning   [code]
F. Benzing*, M. Gauy*, A. Mujika, A. Martinsson and A. Steger
Proceedings of the 36th International Conference on Machine Learning (ICML), 2019

Long Directed Rainbow Cycles and Rainbow Spanning Trees
F. Benzing and A. Pokrovskiy and B. Sudakov
European Journal on Combinatorics, 2020