New preprint: Uncertainty estimation with prediction error circuits
December 15, 2023
Check out our new preprint on Knowing what you don’t know: Estimating the uncertainty of feedforward and feedback inputs with prediction-error circuits!
New preprint: Transcriptomic correlates of state modulation
December 6, 2023
Check out our new preprint on Transcriptomic correlates of state modulation!
Bernstein conference: posters and workshops
September 26, 2023
This week, the Bernstein Conference takes place in Berlin. Friedrich, Loreen, Mark and Joram are presenting posters. Joram and (alumnus) Laura are also organizing workshops:
on behavior across the tree of life and
behavioral flexibility and its neural correlates
'Lottery Tickets in Evolutionary Optimization: On Sparse Backpropagation-Free Trainability' Accepted at ICML 2023!
June 3, 2023
In the paper, Rob and Henning investigated the lottery ticket phenomenon in Evolutionary Optimization. Checkout the preprint and the code.
Evo-devo manuscript published!
May 4, 2023
Joram’s and Henning’s manuscript on the evolution and development of interneurons has been published in Frontiers in Neural Circuits!
You can read it here, or have a look at the visual summary on Twitter.
Open 5-year PhD position in Computational Neuroscience
March 17, 2023
If you are interested in working with us, check out our Jobs page!
Sprekeler Lab @ Cosyne 2023
March 9, 2023
Several of us will be attending Cosyne, this week in Montreal! Denis, Joram and Laura will present posters. Laura will also present in the Workshop on top-down
interactions. Loreen will present in the workshop on diverse inhibition.
Collaboration with Letzkus lab published in Neuron
March 1, 2023
Our collaboration led by the lab of Johannes Letzkus is published in Neuron.
In the paper, Anna Schroeder and collegues analyze the role of inhibitory top-down projections in memory formation.
New preprint on the evolution & development of interneurons
February 28, 2023
In a new preprint, we compare the evolution and development of inhibitory interneurons with functional predictions.
You can find a visual summary on Twitter.
SCIoI Collaboration 'Fish shoals resemble a stochastic excitable system driven by environmental perturbations' Published in Nature Physics!
February 7, 2023
Our SCIoI collaboration with the lab of Pawel Romanzcuk was published in Nature Physics.
In the paper Luis Gomez-Nava and his co-authors analyze the collective diving behavior of fish shoals in Mexico.
They find that real-life fish operate close to a phase transition (‘critical point’).
Afterwards, they combine Machine Learning techniques and in-silico simulation to analyze the function of this system behavior.
They demonstrate that it facilitates optimal information propagation in the face of environment perturbations such as predator attacks.
'Discovering Evolution Strategies via Meta-Black-Box Optimization' Accepted at ICLR 2023!
February 1, 2023
Rob Lange’s DeepMind internship paper has been accepted at ICLR 2023. In the paper he proposes to meta-learn attention-based evolution strategies.
The resulting ‘Learned Evolution Strategy’ outperforms baselines on challenging neuroevolution tasks.
Furthermore, it can be meta-trained in a self-referential fashion and reverse engineered into an analytical Evolution Strategy.
Checkout the preprint and the OpenReview discussion.
Laura's and Joram's paper on invariant neural subspaces is now out in eLife!
April 20, 2022
The paper is available on eLife’s website or under our list of publications.
'On Lottery Tickets and Minimal Task Representations in Deep Reinforcement Learning' Spotlight Accept at ICLR 2022!
February 10, 2022
In the paper, Rob Lange and Marc Visher investigated the lottery ticket phenomenon in Deep Reinforcement Learning. Checkout the preprint and the OpenReview discussion.
New preprint with the Letzkus Lab on long range inhibition
February 8, 2022
In the preprint from first author Anna Schoeder (Letzkus Lab, Freiburg University),
we investigate the role of long range inhibition in memory formation.
Check it out here.
Our paper 'Learning not to learn: Nature vs nurture in silico' was accepted for AAAI '22!
December 13, 2021
In the paper, Rob Lange used meta-learning to study under which conditions non-adaptive heuristics are better than (Bayesian) learning. A preprint is available here.
New preprint out! 'Invariant subspaces maintained by feedback modulation' by Laura and Joram.
November 1, 2021
You can access the paper here .
Preprint from Filip: Self-organization of a doubly asynchronous irregular network state for spikes and bursts
June 25, 2021
Check out our new preprint with Richard Naud. The paper can be found here and here is a tweetprint for you.
Owen's paper on learning E/I assemblies is now out in eLife!
April 26, 2021
The paper is available on eLife’s website or under our list of publications.
Our next preprint is out! Systems memory consolidation by Hebbian plasticity in parallel synaptic pathways
December 7, 2020
Check our new preprint with Michiel Remme, Urs Bergmann, Denis Alevi, Susanne
Schreiber and Richard Kempter! Hebbian plasticity in parallel synaptic pathways: A circuit mechanism for systems memory consolidation. And here is a tweetprint for you.
Preprint from Joram: Learning interneuron diversity
November 19, 2020
Check out our new preprint from Joram & Henning showing that Interneuron diversity is required for compartment-specific feedback inhibition
Our modelling contribution to work by Johannes Letzkus' group has been published in Science!
November 13, 2020
We are very grateful that we could cotribute some theoretical insights to the epxerimental findings by Belén Pardi et al. Read up on how top-down thalamo-cortical inputs are involved in associative memory here.
Workshop on "The making and breaking of E/I balance" taking place on September, 29!
September 26, 2020
We will be hosting a Satellite Workshop of the BCCN conference “The making and breaking of E/I balance”!
We are very excited about an amazing line-up of speakers. If you are interested, join us on Tuesday, September 29 (registration is still open and free).
Loreen's paper on prediction error circuits accepted by eLife! Congratulations!
August 17, 2020
The impatient among you can check out the earlier preprint, but you’ll miss some new results!
Laura's paper on presynaptic inhibition appearing in PLoS Computational Biology! Congrats, Laura!
August 16, 2020
Read up here how presynaptic inhibition helps to stabilize recurrent plastic networks.
Preprint #3 v2 from Owen: Learning E-I assemblies with local rules
June 4, 2020
We’ve posted a new version of our paper on learning E/I assemblies, now with local plasticity rules! We now show that in our recurrent network, feedback alignment can replace backprop and has a simple neural interpretation. Check it out here
Preprint #3 from Owen: Learning E-I assemblies
March 31, 2020
Check out our new paper from Owen, Laura & Henning on how to learn excitatory-inhibitory neuronal assemblies in recurrent networks!
Preprint #2 from Loreen: Learning prediction error neurons by balancing excitation and inhibition!
February 28, 2020
Check out Loreen’s work on how predictions errors can be learned in cortical interneuron circuits!
We are presenting our work at Cosyne 2020!
February 25, 2020
Check out some of our work at the following posters and workshop:
Joram Keijser will present his work “Interneuron diversity emerges in spiking networks optimized for a compartment-specific excitation-inhibition balance” (poster II-39)
Loreen Hertäg will present her work “Interneuron circuits shape the development of negative prediction-error neurons by balancing excitation and inhibition” (poster II-106)
Henning Sprekeler will talk about “Short term plasticity and ML-designed synapse types” during the Learning to Learn workshop
Laura's first preprint is out!
February 20, 2020
Check out how presynaptic inhibition can provide stability in recurrent plastic networks.
Three new lab members joining in 2020!
February 19, 2020
We’re very happy that Felix Lundt, Mark Boon and Yigit Öksuz will join our lab over the course of 2020.
Felix will further strengthen our inhibitory circuits team, and Mark and Yigit our Science of Intelligence efforts,
studying architectural design rules and how mice learn from each other.
PhD position - Social learning in mice and machines
November 14, 2019
We just got a positive funding decision for our project on social learning in mice and machines, within the Excellence Cluster Science of Intelligence. If you are interested in joining the perspectives of machine learning and behavioral biology, have a look at our Jobs page!
Research position (PhD or Postdoc) - Architectural design principles in neural networks
September 25, 2019
We have an open position (PhD or Postdoc) in the Excellence Cluster “Science of Intelligence”. If you’re interest in using ANNs to study the relation between network architecture and task demands, have a look at our Jobs page!
Open postdoc position in Computational Neuroscience
August 25, 2019
If you are keen on understanding the functional role of synaptic plasticity and inhibitory circuits, have a look at our Jobs page!
Robert Lange joins the lab as a PhD student. Welcome!
June 3, 2019
Robert will do his PhD investigating how we can use machine learning tools to learn intelligent collective behaviors, as part of the Einstein Center for Neuroscience and the Cluster of Excellence Science of Intelligence.
Denis Alevi is starting his PhD in the lab!
March 29, 2019
Denis will be working on a mechanistic theory of systems memory consolidation as part of the collaborative research center SFB1315: Mechanisms and disturbances in memory consolidation: from synapses to systems.
New visiting PhD student: Christopher Currin on spiking Epilepsy!
September 23, 2018
Chris is doing his PhD in the Raimondo lab in Cape Town, South Africa, where he
is working on Epilepsy. He is visiting the lab for a 3-months stay to develop a
spiking neural network model and fit it to experimental Epilepsy data from the
Raimondo lab. Welcome, Chris!
