no code implementations • ICLR 2019 • Bernd Illing, Wulfram Gerstner, Johanni Brea
An appealing alternative to training deep neural networks is to use one or a few hidden layers with fixed random weights or trained with an unsupervised, local learning rule and train a single readout layer with a supervised, local learning rule.
1 code implementation • NeurIPS 2023 • Berfin Şimşek, Amire Bendjeddou, Wulfram Gerstner, Johanni Brea
Approximating $f^*$ with a neural network with $n< k$ neurons can thus be seen as fitting an under-parameterized "student" network with $n$ neurons to a "teacher" network with $k$ neurons.
no code implementations • 25 Apr 2023 • Flavio Martinelli, Berfin Simsek, Wulfram Gerstner, Johanni Brea
Can we identify the parameters of a neural network by probing its input-output mapping?
no code implementations • 9 Mar 2023 • Valentin Schmutz, Johanni Brea, Wulfram Gerstner
Can the dynamics of Spiking Neural Networks (SNNs) approximate the dynamics of Recurrent Neural Networks (RNNs)?
2 code implementations • 25 Jan 2023 • Johanni Brea, Flavio Martinelli, Berfin Şimşek, Wulfram Gerstner
MLPGradientFlow is a software package to solve numerically the gradient flow differential equation $\dot \theta = -\nabla \mathcal L(\theta; \mathcal D)$, where $\theta$ are the parameters of a multi-layer perceptron, $\mathcal D$ is some data set, and $\nabla \mathcal L$ is the gradient of a loss function.
no code implementations • 2 Sep 2022 • Alireza Modirshanechi, Johanni Brea, Wulfram Gerstner
Going beyond this technical analysis, we propose a taxonomy of surprise definitions and classify them into four conceptual categories based on the quantity they measure: (i) 'prediction surprise' measures a mismatch between a prediction and an observation; (ii) 'change-point detection surprise' measures the probability of a change in the environment; (iii) 'confidence-corrected surprise' explicitly accounts for the effect of confidence; and (iv) 'information gain surprise' measures the belief-update upon a new observation.
no code implementations • 19 Aug 2022 • Georgios Iatropoulos, Johanni Brea, Wulfram Gerstner
We consider the problem of training a neural network to store a set of patterns with maximal noise robustness.
no code implementations • 12 Feb 2022 • Luca Viano, Johanni Brea
Abstract object properties and their relations are deeply rooted in human common sense, allowing people to predict the dynamics of the world even in situations that are novel but governed by familiar laws of physics.
Common Sense Reasoning Model-based Reinforcement Learning +3
1 code implementation • NeurIPS 2021 • Guillaume Bellec, Shuqi Wang, Alireza Modirshanechi, Johanni Brea, Wulfram Gerstner
Fitting network models to neural activity is an important tool in neuroscience.
1 code implementation • 25 May 2021 • Berfin Şimşek, François Ged, Arthur Jacot, Francesco Spadaro, Clément Hongler, Wulfram Gerstner, Johanni Brea
For a two-layer overparameterized network of width $ r^*+ h =: m $ we explicitly describe the manifold of global minima: it consists of $ T(r^*, m) $ affine subspaces of dimension at least $ h $ that are connected to one another.
no code implementations • 25 Sep 2019 • Berfin Simsek, Johanni Brea, Bernd Illing, Wulfram Gerstner
In a network of $d-1$ hidden layers with $n_k$ neurons in layers $k = 1, \ldots, d$, we construct continuous paths between equivalent global minima that lead through a `permutation point' where the input and output weight vectors of two neurons in the same hidden layer $k$ collide and interchange.
no code implementations • 5 Jul 2019 • Johanni Brea, Berfin Simsek, Bernd Illing, Wulfram Gerstner
The permutation symmetry of neurons in each layer of a deep neural network gives rise not only to multiple equivalent global minima of the loss function, but also to first-order saddle points located on the path between the global minima.
no code implementations • 5 Jul 2019 • Vasiliki Liakoni, Alireza Modirshanechi, Wulfram Gerstner, Johanni Brea
Surprise-based learning allows agents to rapidly adapt to non-stationary stochastic environments characterized by sudden changes.
1 code implementation • 27 Feb 2019 • Bernd Illing, Wulfram Gerstner, Johanni Brea
These spiking models achieve > 98. 2% test accuracy on MNIST, which is close to the performance of rate networks with one hidden layer trained with backpropagation.
3 code implementations • 21 Dec 2018 • Jamie Fairbrother, Christopher Nemeth, Maxime Rischard, Johanni Brea, Thomas Pinder
Gaussian processes are a class of flexible nonparametric Bayesian tools that are widely used across the sciences, and in industry, to model complex data sources.
no code implementations • 17 Dec 2018 • Florian Colombo, Johanni Brea, Wulfram Gerstner
As deep learning advances, algorithms of music composition increase in performance.
1 code implementation • ICML 2018 • Dane Corneil, Wulfram Gerstner, Johanni Brea
Modern reinforcement learning algorithms reach super-human performance on many board and video games, but they are sample inefficient, i. e. they typically require significantly more playing experience than humans to reach an equal performance level.
1 code implementation • 20 Nov 2017 • Johanni Brea
Episodic control has been proposed as a third approach to reinforcement learning, besides model-free and model-based control, by analogy with the three types of human memory.
no code implementations • 9 Dec 2016 • Thomas Mesnard, Wulfram Gerstner, Johanni Brea
In machine learning, error back-propagation in multi-layer neural networks (deep learning) has been impressively successful in supervised and reinforcement learning tasks.
no code implementations • 23 Jun 2016 • Florian Colombo, Samuel P. Muscinelli, Alexander Seeholzer, Johanni Brea, Wulfram Gerstner
A big challenge in algorithmic composition is to devise a model that is both easily trainable and able to reproduce the long-range temporal dependencies typical of music.
no code implementations • NeurIPS 2011 • Johanni Brea, Walter Senn, Jean-Pascal Pfister
We consider a statistical framework in which recurrent networks of spiking neurons learn to generate spatio-temporal spike patterns.