Gundlagenforschung schafft Wissen – im Falle des ESI zu den grundlegenden Fragen, wie das Gehirn in Gesundheit und Krankheit funktioniert. Jeder Erkenntnisgewinn ermöglicht ein tieferes biologisches Verständnis der Organfunktion, bedeutet aber zum Beispiel auch Chancen für Anwendungen der Künstlichen Intelligenz und die Weiterentwicklung der Diagnostik und Therapie in der Psychiatrie und Neurologie.

Dabei ist es von zentraler Bedeutung, dass die gewonnenen Erkenntnisse auch die richtigen Personen erreichen und inspirieren, damit daraus neue Anwendungen entstehen.

Konferenzen

Mit der ESI Systems Neuroscience Conference (ESI SyNC) bietet das ESI ein regelmäßiges, anerkanntes fachliches Forum zu wichtigen Forschungsschwerpunkten der systemischen Neurowissenschaften. Für zwei Tage verwandelt sich das Institut mit Vorträgen erstklassiger Speaker, Poster Sessions und intensiven Panel Diskussionen in ein Wissens-Hub mit Brückenschlag zu weiterführenden Forschungsgebieten, wie zum Beispiel der Psychiatrie.

Vorträge

Im Rahmen regelmäßiger Fachvorträge, wie den ESI PhD-PostDoc-Series und den ESI-Lectures, berichten ESI-Forschende und Speaker externer Forschungseinrichtungen über ihre Projekte. Erste Ergebnisse und neueste Erkenntnisse erreichen so unmittelbar die relevante Fachkreise und stoßen damit den direkten, übergreifenden Wissensaustausch an.

ESI PhD-PostDoc-Series

Exploring the functional organisation of category-selective cortex in visual perception and imagery

30. Juni 2026 Jessica Teed (Rademaker Lab)

Have a Mice Trip! Comprehensive Behavioral Profiling of Psychedelic States in Mice

23. Juni 2026 Ana Cukic (ZeroNoise Lab)

The use of sensory cortex for processing both external & internal stimuli

5. Mai 2026 Rosanne Rademaker (Rademaker Lab)

Neural Dynamics in Macaque Area V1 and V4 during a VR Foraging Task

24. März 2026 Shivangi Patel (ZeroNoise Lab)

Revealing dynamics of dendritic integrations through time-delay embedding and transfer entropy

3. März 2026 Enxhela Vardhami (Cuntz Lab)

Towards systems neuroscience in marmosets: first recordings in our freely moving setup

17. Februar 2026 Francesca Lanzarini (Laurens Lab)

Dynamics of naturalistic attention across species

3. Februar 2026 Mina Glukhova (ZeroNoise Lab)

Would You Agree If N Is Three? On Statistical Inference for Small N

27. Januar 2026 Eleni Psarou (ZeroNoise Lab)

Animal pilot studies should not be used to estimate sample size if effect size and population variance are unknown

27. Januar 2026 Alex Bird (Cuntz Lab)

The statistical power of three monkeys

27. Januar 2026 Jean Laurens (Laurens Lab)

Out of sight, but not out of mind: spatiotemporal representations of occluded motion in Early Visual Cortex

20. Januar 2026 Giuliana Giorjiani (Rademaker Lab)

Towards a neuro-behavioral dictionary of psychedelic states in mice

9. Dezember 2025 Muad Abd El Hay (ZeroNoise Lab)

Similarity Induced Memory Bias in Visual Working Memory

18. November 2025 Nursima Ünver Aydingül (Rademaker Lab)

Closing the Loop - from CA1 to MEC

11. November 2025 Maria Corteze (Schmidt Lab)

Data-Driven Interpretable Dynamical Models for Psychiatry

4. November 2025 Manuel Brenner (Brenner Lab)

Data-Driven Interpretable Dynamical Models for Psychiatry

4. November 2025 Manuel Brenner (Brenner Lab)

A dendritic growth rule for modeling the development of single neurons

7. Oktober 2025 Ryan Rahy (Cuntz Lab)

Code reversal between stimulus processing and fading memories in primate V1

1. Juli 2025 Michael Wolff (Rademaker Lab)

Head Direction Cell in the Naturally Behaving Rats

24. Juni 2025 Farzad Nezhad (Laurens Lab)

Mitigating the Prediction-Undertstanding Trade-off in Neural Modeling

17. Juni 2025 Amr Farahat

Modeling human neurodevelopment in vitro using stem cell-delivered neural organoids

10. Juni 2025 Denise Haslinger (Organoid Lab)

Auditory object formation in temporally complex acoustic scenes

27. Mai 2025 Berfin Bastug

How are working memory representations maintained in visual cortex?

29. April 2025 Noa Krause (Rademaker Lab)

Neural embeddings in early visual cortex

22. April 2025 Jahangir Esfandiari

People are sensitive to their uniquely patterned retinal input

1. April 2025 Amit Rawal (Rademaker Lab)

Studying Neuronal Oscillations in the Primate Brain: From Simplified to Naturalistic Stimuli

25. März 2025 Eleni Psarou (ZeroNoise Lab)

What can optimality tell us about neuronal structure and function?

11. März 2025 Alexander Bird (Cuntz Lab)

Early visual cortex is recruited to act as a comparison circuit between mental representations and visual inputs

24. Februar 2025 Maria Servetnik (Rademaker Lab)

An analog-electric implementation of a harmonic oscillator recurrent neural network

18. Februar 2025 Pedro Carvalho

On the functional role of oscillatory dynamics in neocortical circuits

12. Februar 2025 Felix Effenberger

Mechanisms of local and large-scale integration

21. Januar 2025 Martin Vinck

Bats got rhythm: neural oscillations for vocalization and hearing

14. Januar 2025 Julio Hechavarria (Organoid Lab)

Math, Electrodes, and Rotating Chairs: Unspinning the Vestibular System

17. Dezember 2024 Jean Laurens (Laurens Lab)

Catching cognition in the act: New ways to track naturalistic cognitive processing in monkeys and mice

10. Dezember 2024 Martha Havenith & Marieke Schoelvinck (ZeroNoise Lab)

Analysing and modelling dendrites with the TREES Toolbox

26. November 2024 Hermann Cuntz (Cuntz Lab)

Oscillations, waves, and interference: The "lingua franca" of cortical computations

19. November 2024 Wolf Singer

Connectomics of navigation

12. November 2024 Helene Schmidt (Schmidt Lab)

ESI lectures

Neural Representation and Transformation in Human Memory.

30. Juni 2026 Sue-Hyun Lee (Seoul National University)

Memory retrieval and working memory enable the use of stored information to guide ongoing behavior. This requires not only maintaining information in memory but also representing and transforming it according to current behavioral demands. In this talk, I will present two lines of research that address these aspects of human memory. First, I will introduce our work on working memory, demonstrating that object information can be represented in multiple neural formats beyond its original sensory modality. Using functional magnetic resonance imaging (fMRI), we identified supramodal and cross-modal representations that support the flexible maintenance of information across different sensory modalities. Second, I will present our recent work on goal-directed memory retrieval. By tracking the temporal evolution of neural representations during retrieval, we found that retrieval begins with transient perception-like representations, followed by representational reconfiguration that gives rise to temporally consistent, retrieval cue-specific representations. These findings suggest that memory retrieval involves an active transformation of neural representations rather than a simple reinstatement of stored information. Together, these findings suggest that memory representations are not fixed entities but adaptive neural codes that are organized and reorganized to meet current behavioral demands.

Efficient value processing through primate vision and touch

30. Juni 2026 Hyoung Kim (Seoul National University)

Primates have neural circuits that support efficient and adaptive decision-making. This seminar focuses on two primate-specific features: cognitive flexibility mediated by vision and value-based decision-
making driven by fingertip touch. Cognitive flexibility relies on rostral basal ganglia circuits, while the caudal basal ganglia support habitual stability. These systems operate in parallel, enabling flexible adaptation and stable behavior through efficient neural resource allocation. In addition, primates use fingertip touch for object evaluation and choice. I present that the putamen neurons encode tactile value and integrate visual and tactile information, enabling efficient value abstraction. Together, these mechanisms illustrate how the primate brain achieves efficient value processing and provide insights for
brain-inspired artificial intelligence and treatments for brain disorders.

Bringing psychological and neuroscience perspectives to clinical neurology

22. Juni 2026 Chaipat Chunharas (Chulalongkorn University, Bangkok, Thailand)

The Hippocampal Latent Diffusion Engine: A Computational Framework for Memory, Perception, and Cognitive Dysfunction

8. Juni 2026 Prof. Michael Breakspear (University of Newcastle, Australia)

Drawing on recent advances in generative artificial intelligence and systems neuroscience, we conceptualize the hippocampus as a latent diffusion engine that compresses sensory, internal, and cognitive inputs into low-dimensional representations and then regenerates percepts, episodic memories, and imagined scenes for cortical integration. Travelling waves of cortico-hippocampal oscillations, aligned with large-scale functional gradients, schedule and structure this generative process, organizing distributed neural activity into coherent, semantically grounded constructs. We also demonstrate how major neurodegenerative dementias - including the Alzheimer’s disease spectrum - can be interpreted as selective breakdowns of the anatomical and computational components of this proposed architecture.

Learning to see Stuff

29. Mai 2026 Roland W. Fleming (Experimental Psychology, Justus Liebig University, Giessen)

Every time we open our eyes, we gain access to a rich world of visual experiences. Without having to reach out and touch our surroundings, we can make detailed judgments about the physical properties of objects and materials. How does the brain achieve this, given the inherent ambiguity of raw sensory signals? I’ll present research from my lab on how we see, imagine, understand and interact with objects and materials. I’ll discuss how the visual system estimates material properties like glossiness and viscosity, and how we've been tapping into the underlying perceptual and neural computations. I’ll also suggest that material perception poses fundamental challenges for widely-held tenets of sensory neuroscience, like inverse optics. In addition to the 'perceptual inference' problem of estimating what's 'out there' for any given
image, there's an even more fundamental 'ontological inference' challenge of working out what latent variables need to be estimated in the first place. I'll show how unsupervised learning helps the brain overcome this challenge, allowing us to predict both the broad successes of perception as well as specific illusions.

Hallucinating mice, dopamine and immunity - towards mechanistic treatments for psychosis

24. April 2026 Katharina Schmack (Francis Crick Institute of London)

Psychosis, defined by disturbances of perception and thought, is a core feature of severe mental illness, yet no mechanistic treatments are currently available. A major barrier to progress is the difficulty of modelling psychosis in animals. In our research, we develop and exploit cross-species frameworks for the mechanistic study of psychosis. To measure psychosis phenotypes in mice, we developed a cross-species behavioural–computational paradigm that quantifies hallucination-like perception and reveals how striatal dopamine and acetylcholine shape perceptual inference. To model psychosis-like states, we established a mouse model of autoimmune psychosis demonstrating that brain-reactive autoantibodies can drive psychosis-like behaviour, and how antipsychotic drugs may act through immunomodulatory mechanisms. Together, these findings point toward new biologically grounded strategies for treating psychosis.Abstract will follow shortly.

Functional hyperconnectivity related to brain disease: maladaptive process or element of resilience?

14. April 2026 Markus Aswendt (Department of Experimental Neurology, KGU)

Stroke affects nearly 12 million people worldwide each year and remains a leading cause of long-term disability. Recovery depends on neural plasticity processes that reorganize brain circuits after injury, which can be assessed longitudinally using resting-state functional MRI (rs-fMRI). While early functional hyperconnectivity has often been interpreted as a compensatory response, growing evidence across neurological disorders suggests that persistent hyperconnectivity may represent a maladaptive network state associated with inefficient processing and impaired functional recovery1 .

To investigate functional connectivity dynamics during stroke recovery, we established resting-state fMRI in mice together with automated post-processing pipelines for standardized and reproducible analysis2,3 . Mice subjected to focal ischemic stroke received daily intranasal treatment with the complement peptide C3a, a signaling molecule known to enhance neural plasticity4 or vehicle control for three weeks. Early after stroke, we observed hypoconnectivity in stroke-affected brain regions. Control-treated mice later developed widespread hyperconnectivity, whereas C3a-treated mice showed connectivity patterns largely restored to pre-stroke levels. Importantly, increased connectivity in stroke-unaffected regions was associated with poorer motor recovery.

Cortical dynamics for adaptive computation in biological and artificial networks

23. März 2026 Nuttida Rungrasameetaweemana (Columbia University)

The ability to flexibly use prior knowledge to guide behavior in changing sensory environments is a hallmark of natural intelligence. Yet how the brain supports adaptive computation, and which components of this process break down in neuropsychiatric and neurological disorders, remains poorly understood. In this talk, I will share a series of studies from our group that combine human single-neuron electrophysiology, neuroimaging, lesion approaches, and biologically grounded network models to examine adaptive computation across levels, from individual neurons and distributed circuits to emergent behavior. I will also discuss which computational principles may generalize across biological and artificial networks.

Working memory beyond the cortex

20. März 2026 Anastasia Kiyonaga (Department Of Cognitive Science, University of California, USA)

The core role of working memory (WM) is to sustain short-term goal content in mind for long enough to guide behavior. Yet the nervous system may engage myriad functions to accomplish that feat. WM signals are widely distributed across the brain and can now be detected in evolutionarily earlier structures than previously thought—like the cerebellum, thalamus, and superior colliculus. Mounting evidence also shows that even peripheral oculomotor and physiological signals can carry WM feature content, raising the question of just how far WM signals extend. I will describe a recent body of work where we test the idea that visual WM content is adaptively distributed across the nervous system—including the sensory receptors and motor effectors—according to behavioral demands. For instance, we find feature-specific WM content evidence in signatures such as pupil size modulations, spatio-temporal gaze sequences, and systematic manual gestures. We also find that such signatures are tailored to the unique behavioral demands of the situation, hinting that they are selective and functionally relevant. These findings converge on the idea that WM may functionally recruit even the most primary structures in sensorimotor processing, to mentally represent goals and turn them into actions.

Rapid dynamics and long-term drift in working memory

10. März 2026 John Serences (Department of Psychology and Neuroscience Graduate Program, University of California, USA)

Working memory enables the temporary storage and manipulation of information to guide goal‑directed behavior. Traditional models posit that working memory is supported by sustaining sensory-evoked patterns of neural activity after a stimulus is no longer present in the environment. However, recent work demonstrates that neural codes related to working memory rapidly change over a short timescale of seconds (short-term drift) and even over the course of days/months, well after task performance has stabilized (long-term representational drift). Short‑term drift is thought to reflect a rotation of stimulus‑specific neural codes that helps shield memory representations from interference due to new sensory inputs. Long‑term drift has been linked to changes in coding sparsity and efficiency during consolidation, though some accounts suggest it may instead reflect a random walk that occurs in any high‑dimensional system that exhibits plasticity.

To examine mechanisms underlying short‑term drift, we trained continuous‑time recurrent neural networks (CTRNNs) on a working memory task that incorporated distractor interference and variable stimulus-response mappings. Contrary to theories emphasizing protective rotations of memory representations, our results indicate that short‑term dynamics primarily arise from the mixing of signals related to stimulus storage, decision formation, and response preparation. To test accounts of long-term drift, we used fMRI to collect data from the same subjects repeatedly over ~60-180 days while they performed a working memory task. We find that activation patterns related to memory load exhibit directional drift, paralleling observations made using other tasks in animal model systems. Complementary work using CTRNNs showed that representational drift naturally emerges when networks continue to train after reaching stable task performance. Notably, this drift is accompanied by increasingly sparse neural codes that are more resilient to distractor interference and other task perturbations. Together, these results suggest that short‑term drift reflects interactions between multiple cognitive processes rather than a mechanism solely dedicated to preventing distractor interference, and that long‑term drift reflects an ongoing refinement of neural codes that enhances the efficiency and robustness of working memory representations.

Cross-species study of statistical learning, with and without feedback

20. Februar 2026 Athena Akrami (Sainsbury Wellcome Centre, UCL)

Refinement and Reduction Strategies

19. Dezember 2025 Univ.-Prof. Dr. med.vet Christa Thöne-Reineke (Institut für Tierschutz, Tierverhalten und Versuchstierkunde)

With European Directive 20210/63, the internationally recognised 3R principle (Replace, Reduce, Refine) was incorporated into European and national legislation. The implications of this directive for applications involving animal experimentation are briefly outlined. Using project examples from the School of Veterinary Medicine at Freie Universität Berlin, refinement strategies in teaching and research are presented, focusing on stress assessment from the perspective of mice, alternative methods in teaching, cognitive enrichment and training, husbandry, and pain management. Finally, the fate of experimental animals at the conclusion of studies and the establishment of a culture of care are addressed. The outlook introduces the 3REthicWeb and GRK3051-3RTG projects.

Emergence of structure and function through synaptic plasticity

24. Oktober 2025 Julijana Gjorgjieva (TU München)

Synapses and neuronal properties in neural circuits are constantly modified by various plasticity mechanisms operating at different timescales. While extensive experimental studies have characterized these plasticity mechanisms, understanding their functional implications is challenging due to their interplay, and the complexity induced by the diversity of cell types. Combining theoretical analysis and computational modeling, I will discuss how different plasticity mechanisms refine connectivity giving rise to structures such as E-I co-tuning and lateral inhibition. I will show what these different connectivity structures imply in terms of computations, and how they can be further modified by contextual signals.

Controlling the Contents of Working Memory

29. August 2025 Brad Postle (University of Wisconsin-Madison)

Working memory poses a problem of stability versus flexibility for the cognitive system. I will present work that operationalizes this problem as one of prioritization versus active removal: In some situations the brain must differently process and represent the information that is needed to guide immediate action versus the information that is not needed “now,” but may be needed “later”; in others, no-longer-relevant information must be discarded to free up capacity and minimize potentially deleterious interference. Research with fMRI, EEG, and recurrent neural network (RNN) simulation is revealing, at the levels of algorithmic mechanism and neural implementation, how the human brain accomplishes these aspects of the control of the contents of working memory.

Real world neuroscience

23. Juni 2025 Saikat Ray (Ulanovsky Lab, Weizmann Institute of Science)

How does the brain actually represent the real world – in natural and multi-animal settings? I will describe my work with Egyptian fruit bats, where I performed wireless neural recordings in the hippocampal formation in freely flying bats – in social groups in a laboratory-based cave, and outside on an oceanic island. These studies provide the first insights into the neural encoding of real-world places and societies.

Microglia, their shape, and their impact on neuronal function after psychedelic exposure

21. Februar 2025 Sandra Siegert (Institute of Science and Technology Austria (ISTA), Austria)

Brain dysfunction has a detrimental effect on our daily lives. Yet, surprisingly little is known about how microglia, traditionally classified as immune-responsive cells embedded within the neuronal network, maintain brain functionality and how deviation from their tasks will induce disease onset.
My talk will start with a short trajectory on how I shifted my neuro-centric perspective of the brain toward microglia. Then, I will present our initial struggles using the oversimplified assumption and methodologies in the glia field to classify microglia morphology. I will introduce our newly developed classifier morphOMICs to overcome these limitations. Finally, I will show how psychedelics like ketamine in the adult brain enable microglia to adapt the extracellular space in a sex-specific manner and, in this way, influence new synapse formation.

Martina Pandinelli moderiert ESI Sync – Ernst Struengmann Institute for Neuroscience ESI Frankfurt

Die Konferenz wird von ESI-Nachwuchswissenschaftler:innen konzipiert, geplant, organisiert und durchgeführt.

Wissenschaftler erläutert Forschungsergebnisse während Postersession – Ernst Struengmann Institute for Neuroscience ESI Frankfurt

Im Mittelpunkt steht der Austausch zu richtungweisenden Erkenntnissen.

Publikum bei ESI SyNC – Ernst Struengmann Institute for Neuroscience ESI Frankfurt

Teilnehmende interdisziplinärer Fachbereiche profitieren unmittelbar von den Informationen.

Outreach

Die gewonnenen Erkenntnisse breit mit der interessierten Öffentlichkeit zu teilen, Fragen zur Forschung zu beantworten und einen Beitrag dazu zu leisten, den Wert der Grundlagenforschung zu vermitteln, ist ein besonderes Anliegen des Instituts. Dazu tragen vielseitige, individuell zugeschnittene Veranstaltungsformate, wie zum Beispiel regelmäßige Aktionstage anlässlich des Girls’Day oder die Teilnahme des ESI am Science Festival der Frankfurt Alliance bei.

Offene Positionen

Alle offenen Positionen ansehen

Vielen Dank für Ihr Interesse, am ESI mitzuarbeiten. Wir freuen uns auf ihre Bewerbung als …

PhD Kandidat:innen

Postdoc Kandidat:innen

Gedanken anregen, Wissen teilen