Biology subjects

Woergoetter, F.

Publications and source records attributed to Woergoetter, F..

6 recordsLinked to original sources

bioRxiv2026

Meta-learning leading to homeostatic plasticity stabilizes synaptic weights together with predictable activity levels

Stabilizing synaptic plasticity together with the neurons activity has remained a central challenge in theoretical neuroscience since the introduction of Hebbian learning principles. Classical Hebbian learning rules typically lead to unbounded synaptic growth, motivating the development of stabilization mechanisms such as normalization methods, BCM-type learning, synaptic scaling and others. While these approaches can prevent divergence, they can also exhibit different limitations e.g. resulting in too-sparse synaptic configurations or leading to poor scalability with increasing network size. A recently introduced meta-plasticity mechanism, termed annealed linear learning (ALL), dynamically reduces the learning rate as neuronal output increases, thereby preserving stable and interpretable fixed-point behavior of the output. However, the original formulation leads to an irreversible decay of the learning rate, preventing adaptation to changing environmental conditions. To address this, in the present study, we balance learning rate reduction at large outputs with recovery at small outputs and in addition introduce forgetting that gradually reduces synaptic weights. These extensions allow the system to discard outdated representations and adapt to novel input conditions. Analytical investigations demonstrate that the favorable output fixed-point properties of the original ALL framework are preserved under the extended rule. Furthermore, simulations with an artificial agent show that the proposed mechanism enables robust and fast re-learning and adaptation in changing environments.

neuroscience↗

bioRxiv2025

Analyzing Gaze and Hand Movement Patterns in Leader-Follower Interactions During a Time-Continuous Cooperative Manipulation Task

In daily life, people often interact by taking on leader and follower roles. Unlike laboratory experiments, these interactions unfold naturally and continuously. Although it is well established that gaze typically precedes object manipulation, much less is known about how gaze-hand patterns evolve in interactive settings where one person must take the others actions into account. Here we examine predictive, planning-related behavior in a two-player tabletop game called "do-undo." Participants alternated as Leader and Follower. The Leader performed simple pick-and-place actions to alter the arrangement of objects, while the Follower used other objects to restore the previous configuration. We recorded eye and hand movements, along with object trajectories, using a system that combined eye tracking with multi-camera motion capture. Touch sensors on the players hands provided precise timing of contacts, allowing us to segment cooperative action into well-defined temporal intervals. As expected, eye fixations consistently preceded manipulation, but clear role differences emerged. Leaders looked more often and earlier at target objects. In many trials, their gaze anticipated not only their own actions but also those required of the Follower. Leaders also more frequently checked the outcome of the do-undo sequence. Both roles showed gaze patterns consistent with memorization, but alternating gazes between objects and destinations were much more common in Leaders. Some patterns suggested longer-term planning beyond the immediate action. These findings reveal distinct decision-making and planning strategies in Leaders and Followers. Leaders consider not only their own next moves but also the potential actions of their partners, shedding light on the complex cognitive processes that underly everyday human interaction.

neuroscience↗

bioRxiv2025

Human Action Planning and Mental Rotation in a Tetris-Like-Game

The mechanisms behind human action planning and mental object manipulation are still not well understood. These core cognitive abilities are essential not only for interaction with physical spaces, e.g., for assembling objects, but also for effective problem-solving in the digital world. Here we ask, which strategies humans employ when assessing whether or not an object will fit into a cavity. To this end, objects and cavities were presented with different orientations on a computer screen and we measured errors, reaction-times and gaze patterns, where the latter can point to different problem-solving strategies. On the one hand our findings confirm that simpler configurations are solved faster and more efficiently than more complex ones. On the other hand, by analyzing about 80,000 gazes, we observed that participants used three different strategies. In many instances, the investigated task -- featuring relatively large objects -- could be completed using only peripheral vision (37%). In a larger number of cases quite "specific" gaze patterns were observed, primarily focusing on the Gestalt of a concave corner (46%). Less frequently, but still notably often, participants employed a strategy of fixating near to object or cavity (17%), potentially minimizing the length of the required saccadic eye movements while relying on perifoveal/peripheral vision. Ultimately, these findings highlight the crucial roles of proximity, spatial orientation, and visual cues in object recognition tasks, suggesting that the perceptual strategies used depend on distinct aspects of the object configurations.

physiology↗

bioRxiv2023

Unsupervised Learning of Perceptual Feature Combinations

In many situations it is behaviorally relevant for an animal to respond to co-occurrences of perceptual, possibly polymodal features, while these features alone may have no importance. Thus, it is crucial for animals to learn such feature combinations in spite of the fact that they may occur with variable intensity and occurrence frequency. Here, we present a novel unsupervised learning mechanism that is largely independent of these contingencies and allows neurons in a network to achieve specificity for different feature combinations. This is achieved by a novel correlation-based (Hebbian) learning rule which allows for linear weight growth and which is combined with a mechanism for gradually reducing the learning rate as soon as the neurons response becomes feature combination specific. In a set of control experiments, we show that other existing advanced learning rules cannot satisfactorily form such multi-feature representations. In addition, we show that networks, which use this type of learning always stabilize and converge to subsets of neurons with different feature-combination specificity. Neurons with this property may, thus, serve as an initial stage for the processing of ecologically relevant real world situations for an animal.

neuroscience↗

bioRxiv2020

Exploring new roles for actin upon LTP induction in dendritic spines

Dendritic spines, small protrusions of the dendrites, enlarge upon LTP induction, linking morphological and functional properties. Although the role of actin in spine enlargement has been well studied, little is known about its relationship with mechanical membrane properties, such as membrane tension, which is involved in many cell processes, like exocytosis. Here, we use a 3D model of the dendritic spine to investigate how polymerization of actin filaments can effectively elevate the membrane tension to trigger exocytosis in a domain close to the tip of the spine. Moreover, we show that the same pool of actin promotes full membrane fusion after exocytosis and spine stabilization.

neuroscience↗

bioRxiv2020

Actin in Dendritic Spines Self-Organizes into a Critical State

It is known that dendritic spines change their size and shape spontaneously and sometimes to a large degree, but the function of this remains unclear. Here, we quantify these changes using time-series analysis of confocal data and demonstrate that spine size can follow different autoregressive integrated moving average (ARIMA) models and that shape- and size-changes are not correlated. We capture this behavior with a biophysical model, based on the spines actin dynamics, and find the presence of 1/f noise. When investigating its origins, the model predicts that actin in the dendritic spines self-organizes into a critical state, which creates a fine balance between static actin filaments and free monomers. We speculate that such a balance might be functionally beneficially to allow a spine to quickly reconfigure itself after LTP induction.

biophysics↗