I’m a co-founder of twig.energy, a startup where we build software enabling the zero-carbon transition of the power grid (If you’re interested in learning more feel free to reach out!). Previously I was a ML/DL research scientist at Google Brain where I started and led Google Research’s efforts within data driven weather forecasting. I am broadly interested in both theoretical and applied machine learning and have extensive experience in deep learning, probabilistic inference and large scale distributed learning. Particularly I’m interested in applying Machine learning to solve important problems within climate mitigation, bioinformatics and energy. I have published papers on major machine learning conferences such as NeurIPS, ICLR and ICML and scientific journals such as Nature Bioinformatics and Oxford bioinformatics. I completed a Phd degree at the University of Copenhagen supervised by professor Ole Winther where I also spend time with DeepMind technologies and Twitter Cortex VX in London.

casperkaae@gmail.com Google Scholar Profile CV GitHub LinkedIn Twitter


Deep learning for twelve hour precipitation forecasts
L. Espeholt, S. Agrawal, C. K. Sønderby, M. Kumar, J. Heek, C. Bromberg, C. Gazen, R. Carver, M. Andrychowicz, J. Hickey, A. Bell, N. Kalchbrenner
Nature Communications, 2022
PDF Abstract Bibtex Blog

Skillful Twelve Hour Precipitation Forecasts using Large Context Neural Networks
L. Espeholt, S. Agrawal, C.K. Sønderby, M. Kumar, J. Heek, C. Bromberg, C. Gazen, J. Hickey, A. Bell, N. Kalchbrenner
Arxiv, 2021
PDF Abstract Bibtex Blog

IDF++: Analyzing and Improving Integer Discrete Flows for Lossless Compression
R. vd Berg, A. Gritsenko, M. Dehghani, C. K. Sønderby, T. Salimans
ICLR, 2020
PDF Abstract Bibtex

MetNet - A Neural Weather Model for Precipitation Forecasting
C.K. Sønderby, L. Espeholt, J. Heek, M. Dehghani, A. Oliver, T. Salimans, S. Agrawal, J. Hickey, N. Kalchbrenner
Arxiv, 2020
PDF Abstract Bibtex Blog

SignalP 5.0 improves signal peptide predictions using deep neural networks
J.J.A Armenteros and K.D Tsirigos, C.K. Sønderby, T.N. Petersen, O. Winther, S. Brunak, G.v Heijne, H. Nielsen
Nature biotechnology, 2019
Abstract Bibtex Website

An introduction to deep learning on biological sequence data - examples and solutions
V.I Jurtz, A.R. Johansen, M. Nielsen, J.J.A Armenteros, H. Nielsen, C.K. Sønderby O. Winther S.K. Sønderby
Oxford Bioinformatics, 2017
Abstract Bibtex Code

Continuous Relaxation Training of Discrete Latent Variable Image Models
C.K. Sønderby, B. Poole, A. Mnih
Bayesian DeepLearning Workshop @ Nips, 2017
PDF Abstract Bibtex

Deep Recurrent Conditional Random Field Network for Protein Secondary Prediction
A.R. Johansen, C.K. Sønderby, S.K. Sønderby, O. Winther
ACM-BCB 2017, 2017
PDF Abstract Bibtex

DeepLoc - Prediction of protein subcellular localization using deep learning
J.A. Armenteros, C.K. Sønderby, S.K. Sønderby, H. Nielsen, O. Winther
Oxford Bioinformatics, 2017
Abstract Bibtex Code Website

Amortised MAP Inference for Image Super-resolution
C.K. Sønderby, J. Caballero, L. Theis, W Shi, F. Huszár
ICLR (Oral presentation), 2017
PDF Abstract Bibtex Blog

Ladder Variational Autoencoders
C.K. Sønderby, T. Raiko, L. Maaløe, S.K. Sønderby, O. Winther
Neural Information Processing Systems (NIPS), 2016
PDF Abstract Bibtex Code

Auxiliary Deep Generative Models
L. Maaløe, C.K. Sønderby, S.K. Sønderby, O. Winther
International Conference on Machine Learning (ICML), 2016
PDF Abstract Bibtex Code

BloodSpot - a database of gene expression profiles and transcriptional programs for healthy and malignant haematopoiesis
F.O. Bagger, D. Sasivarevic, S.H. Sohi, L.G. Laursen, S. Pundhir, C.K. Sønderby, O. Winther, N. Rapin, B.T. Porse
Nucleic acids research, 2016
PDF Abstract Bibtex Website

Convolutional LSTM networks for subcellular localization of proteins
S.K. Sønderby, C.K. Sønderby, H. Nielsen, O. Winther
International Conference on Algorithms for Computational Biology, 2015
PDF Abstract Bibtex

Tumor suppressor ASXL1 is essential for the activation of INK4B expression in response to oncogene activity and anti-proliferative signals
X. Wu, I.H. Bekker-Jensen, J. Christensen, K.D. Rasmussen, S. Sidoli, Y. Qi, Y. Kong, X. Wang, Y. Cui, Z. Xiao, G. Xu, K. Williams, J. Rappsilber, C.K. Sønderby, O. Winther, O.N. Jensen, K. Helin
Cell research, 2015
Abstract Bibtex

Recurrent Spatial Transformer Networks
S.K. Sønderby, C.K. Sønderby, L. Maaløe, O. Winther
ArXiv Preprint, 2015
PDF Abstract Bibtex

Diffusion weighted imaging with circularly polarized oscillating gradients
H. Lundell, C.K. Sønderby, T.B Dyrby
Magnetic resonance in medicine, 2015
Abstract Bibtex

Apparent exchange rate imaging in anisotropic systems
C.K. Sønderby, H. Lundell, L.V. Søgaard, T.B Dyrby
Magnetic resonance in medicine, 2014
Abstract Bibtex

Orientationally invariant metrics of apparent compartment eccentricity from double pulsed field gradient diffusion experiments
S.N. Jespersen, H. Lundell, C.K. Sønderby, T.B Dyrby
NMR in Biomedicine, 2013
Abstract Bibtex


Below you can find some of the material I used for courses and workshops where I have teached.

DTU course 02456 Deep learning
Programming Exercises (Theano/Lasagne) for the Deep Learning Graduate Course at the Technical University of Denmark running in the Fall of 2016

Nvidia Deep Learning Summercamp 2016
Programming Exercises (Theano/Lasagne) for the 2016 Nvidia Deep Learning Summercamp in London.

Deep Learning DTU summer school 2015
Programming Exercises (Theano/Lasagne) for the 2015 Deep Learning Summer school at the Technical University of Denmark. Exercises for low level implementation of a ConvNet in Numpy and high level Theano/Lasagne exercises.