Publications by Johanni Michael Brea

Related publications (17)

Brain signals of a Surprise-Actor-Critic model: Evidence for multiple learning modules in human decision making

Wulfram Gerstner, Johanni Michael Brea, Kerstin Preuschoff, Alireza Modirshanechi, Marco Philipp Lehmann, Vasiliki Liakoni

Learning how to reach a reward over long series of actions is a remarkable capability of humans, and potentially guided by multiple parallel learning modules. Current brain imaging of learning modules is limited by (i) simple experimental paradigms, (ii) e ...

ACADEMIC PRESS INC ELSEVIER SCIENCE2022

Learning in Volatile Environments With the Bayes Factor Surprise

Wulfram Gerstner, Johanni Michael Brea, Alireza Modirshanechi, Vasiliki Liakoni

Surprise-based learning allows agents to rapidly adapt to nonstationary stochastic environments characterized by sudden changes. We show that exact Bayesian inference in a hierarchical model gives rise to a surprise-modulated trade-off between forgetting o ...

MIT PRESS2021

Surprise and novelty in the brain

Wulfram Gerstner, Johanni Michael Brea, Alireza Modirshanechi, Sophia Becker

Notions of surprise and novelty have been used in various experimental and theoretical studies across multiple brain areas and species. However, 'surprise' and 'novelty' refer to different quantities in different studies, which raises concerns about whethe ...

CURRENT BIOLOGY LTD2023

Geometry of the Loss Landscape in Overparameterized Neural Networks: Symmetries and Invariances

Wulfram Gerstner, Clément Hongler, Johanni Michael Brea, Francesco Spadaro, Berfin Simsek

We study how permutation symmetries in overparameterized multi-layer neural networks generate `symmetry-induced' critical points. Assuming a network with

L

layers of minimal widths

r_1^*, \ldots, r_{L-1}^*

reaches a zero-loss minimum at $ r_1^*! \c ...

2021

Biologically plausible deep learning – but how far can we go with shallow networks?

Wulfram Gerstner, Johanni Michael Brea, Bernd Albert Illing

Training deep neural networks with the error backpropagation algorithm is considered implausible from a biological perspective. Numerous recent publications suggest elaborate models for biologically plausible variants of deep learning, typically defining s ...

2019

Auditory stimuli suppress contextual fear responses in safety learning independent of a possible safety meaning

Elena Lidia Jeanine Mombelli, Ralf Schneggenburger, Olexiy Kochubey, Johanni Michael Brea, Denys Osypenko, Shriya Palchaudhuri, Christos Sourmpis

Safety learning allows the identification of non-threatening situations, a learning process instrumental for survival and psychic health. In contrast to fear learning, in which a sensory cue (conditioned stimulus, CS) is temporally linked to a mildly avers ...

2024

On the choice of metric in gradient-based theories of brain function

Wulfram Gerstner, Johanni Michael Brea, Jean-Pascal Théodor Pfister

The idea that the brain functions so as to minimize certain costs pervades theoretical neuroscience. Because a cost function by itself does not predict how the brain finds its minima, additional assumptions about the optimization method need to be made to ...

2020

Efficient Model-Based Deep Reinforcement Learning with Variational State Tabulation

Wulfram Gerstner, Johanni Michael Brea, Dane Sterling Corneil

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. To imp ...

2018

Exponentially long orbits in Hopfield neural networks

Wulfram Gerstner, Johanni Michael Brea, Samuel Pavio Muscinelli

We show that Hopfield neural networks with synchronous dynamics and asymmetric weights admit stable orbits that form sequences of maximal length. For N units, these sequences have length T = 2^N; that is, they cover the full state space. We present a mathe ...

Massachusetts Institute of Technology Press2017

Algorithmic Composition of Melodies with Deep Recurrent Neural Networks

Wulfram Gerstner, Johanni Michael Brea, Samuel Pavio Muscinelli, Alexander Kevin Seeholzer, Florian François Colombo

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. Here we investigate how artificial neural networks can be trained on a large corpus o ...

2016

Decoupling Backpropagation using Constrained Optimization Methods

Martin Jaggi, Johanni Michael Brea

We propose BlockProp, a neural network training algorithm. Unlike backpropagation, it does not rely on direct top-to-bottom propagation of an error signal. Rather, by interpreting backpropagation as a constrained optimization problem we split the neural ne ...

2018

Does computational neuroscience need new synaptic learning paradigms?

Wulfram Gerstner, Johanni Michael Brea

Computational neuroscience is dominated by a few paradigmatic models, but it remains an open question whether the existing modelling frameworks are sufficient to explain observed behavioural phenomena in terms of neural implementation. We take learning and ...

Elsevier Science Bv2016

Kernel Memory Networks: A Unifying Framework for Memory Modeling

Wulfram Gerstner, Johanni Michael Brea, Georgios Iatropoulos

We consider the problem of training a neural network to store a set of patterns with maximal noise robustness. A solution, in terms of optimal weights and state update rules, is derived by training each individual neuron to perform either kernel classifica ...

Curran Associates2022

Computational models of episodic-like memory in food-caching birds

Wulfram Gerstner, Johanni Michael Brea

How the 'what', 'where', and 'when' of past experiences are stored in episodic memories and retrieved for suitable decisions remains unclear. In an effort to address these questions, the authors present computational models of neural networks that behave l ...

Nature Portfolio2023

Testing two competing hypotheses for Eurasian jays' caching for the future

Johanni Michael Brea

Previous research reported that corvids preferentially cache food in a location where no food will be available or cache more of a specific food in a location where this food will not be available. Here, we consider possible explanations for these prospect ...

NATURE RESEARCH2021

Emergent Rate-Based Dynamics in Duplicate-Free Populations of Spiking Neurons

Wulfram Gerstner, Johanni Michael Brea, Valentin Marc Schmutz

Can spiking neural networks (SNNs) approximate the dynamics of recurrent neural networks? Arguments in classical mean-field theory based on laws of large numbers provide a positive answer when each neuron in the network has many “duplicates”, i.e., other n ...

American Physical Society (APS)2025

A taxonomy of surprise definitions

Wulfram Gerstner, Johanni Michael Brea, Alireza Modirshanechi

Surprising events trigger measurable brain activity and influence human behavior by affecting learning, memory, and decision-making. Currently there is, however, no consensus on the definition of surprise. Here we identify 18 mathematical definitions of su ...

2022