Recent searches

No recent searches

Search options

Only available when logged in.
sigmoid.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
A social space for people researching, working with, or just interested in AI!

Administered by:

Server stats:

597
active users

sigmoid.social: AboutProfiles directoryPrivacy policy

Mastodon: AboutGet the appKeyboard shortcutsView source codev4.4.3

#compneuro

3 posts3 participants0 posts today
Fabrizio Musacchio

In our recent , I presented Genkin et al. (2025), who decode in the of as low-dimensional shared across . Their generative model links tuning curves, spike-time variability, and stimulus-dependent potential landscapes to a common internal decision variable. I summarized and discussed their findings in this blog post:

📝doi.org/10.1038/s41586-025-091
🌍fabriziomusacchio.com/blog/202

#CompNeuro#Neuroscience#Cortex
Dan Goodman

New preprint with @marcusghosh on how neural network architecture shapes function. We explored a wide range of architectures, and a family of tasks with components of navigation, decision making under uncertainty, multimodal integration and memory. Performance better explained by "computational traits" like sensitivity and memory, than by architectural features.

biorxiv.org/content/10.1101/20

#neuroscience#compneuro#computationalneuroscience
*
Fabrizio Musacchio

In their study, Morales-Gregorio et al. show that in shift dynamically under top-down influence from . They identify two distinct population activity states – eyes open vs. closed – with notably stronger V4→V1 signaling in the foveal region during eyes-open periods. A cool example of how cognitive context reshapes visual cortical dynamics.

🌍 cell.com/cell-reports/fulltext

#CompNeuro#Neuroscience#VisualCortex
Fabrizio Musacchio

New available: Functional Imaging Data Analysis – From Calcium Imaging to Network Dynamics. This course covers the entire workflow from raw data to functional insights, including & . Designed for students and for self-guided learning, with a focus on open content and reproducibility. Feel free to use and share it 🤗

🌍 fabriziomusacchio.com/blog/202

#Python#DataScience#MachineLearning
Dan Goodman

How can we test theories in neuroscience? Take a variable predicted to be important by the theory. It could fail to be observed because it's represented in some nonlinear, even distributed way. Or it could be observed but not be causal because the network is a reservoir. How can we deal with this?

Increasingly feel like this isn't a theoretical problem but a very practical one that comes up all the time. I'd be interested if anyone has seen anything practical that addresses this.

#neuroscience#compneuro
Alessandro Torcini

#compneuro #neuroscience

Here is the complete list of speakers, title of the talks and the
timeline of the workshop #CNS2025:

doocn.org/

Population activity : the influence of cell-class identity, synaptic
dynamics, plasticity and adaptation

that will take place in Florence 8-9 July 2025 with 16 invited speakers

see you soon
S. Olmi, A. Torcini, M. Giugliano

doocn.orgDOOCN-XVI: Dynamics On and Of Complex NetworksOfficial website of the Dynamics On and Of Complex Networks workshop series
Fabrizio Musacchio

🧠✨ Just published – “The dynamics and geometry of choice in the premotor ” by Genkin et  al. shows how the encodes . Using single-trial , they model as 1D trajectories on a high-dimensional .

Key findings:
👉 Diverse tuning curves reflect one latent decision variable
👉 Dynamics follow an attractor
👉 Geometry links sensory & cognitive coding

✍️ nature.com/articles/s41586-025
💻 github.com/engellab/neuralflow

NatureThe dynamics and geometry of choice in the premotor cortex - NatureA population code for the dynamics of choice formation in the primate premotor cortex is revealed, with diverse single-neuron tuning to a shared decision variable.
#Neuroscience#CompNeuro
Fabrizio Musacchio

🧠 👀 Fascinating new study: Pre-training with spontaneous retinal waves — those endogenous activity patterns in the developing eye — significantly improves motion prediction in natural scenes.

May, Dauphin & Gjorgjieva show that even before , the may self-organize using internally generated signals.

📖 journals.plos.org/ploscompbiol
💻 Code: github.com/comp-neural-circuit

journals.plos.orgPre-training artificial neural networks with spontaneous retinal activity improves motion prediction in natural scenesAuthor summary Before the onset of vision, the retina generates its own spontaneous activity, referred to as retinal waves. This activity is crucial for establishing neural connections and, hence, ensuring the proper functionality of the visual system. Recent research has shown that retinal waves exhibit statistical properties similar to those of natural visual stimuli, such as the optic flow of objects in the environment during forward motion. We investigate whether retinal waves can prepare the visual system for motion processing by pre-training artificial neural network (ANN) models with retinal waves. We tested the ANNs on next-frame prediction tasks, where the model predicts the next frame of a video based on previous frames. Our results showed that ANNs pre-trained with retinal waves exhibit faster learning on movies featuring naturalistic stimuli. Additionally, pre-training with retinal waves refined the receptive fields of ANN neurons, similar to processes seen in biological systems. Our work highlights the importance of spatio-temporally patterned spontaneous activity in preparing the visual system for motion processing in natural scenes.
#Neuroscience#AI#ComputationalBiology
Fabrizio Musacchio

Recently, we discussed this insightful paper by Squadrani et al (2024) in our . It explores how enhance during by modulating D-serine levels, providing a basis for dynamic thresholds. The findings suggest astrocytic signaling is crucial for , linking activity to flexibility. Here’s a summary from our JC:

🌍 fabriziomusacchio.com/blog/202
📝 doi.org/10.1038/s42003-024-065

#CompNeuro#Neuroscience
*
Alicia Izquierdo, Ph.D.

📣 Preprint alert ✨New insights into the tradeoff of effort and delay costs! A collaboration with the Wikenheiser lab #neuroscience #compneuro biorxiv.org/content/10.1101/20

bioRxiv · A progressive ratio task with costly resets reveals adaptive effort-delay tradeoffsThe Progressive Ratio (PR) schedule is a popular test for measuring the motivational value of a reinforcer, in which subjects must exert an increasing amount of work to obtain each successive reward. Despite its popularity, the PR task hinges on a low-dimensional behavioral readout — breakpoint, or the maximum work requirement subjects are willing to complete before abandoning the task. Here, we show that with a simple modification, the PR task can be transformed into an optimization problem reminiscent of the patch-leaving foraging scenario, which has been analyzed extensively by behavioral ecologists, psychologists, and neuroscientists. In the Progressive Ratio with Reset (PRR) task, rats perform the PR task on one lever, but can press a second lever to reset the current ratio requirement back to its lowest value at the cost of enduring a reset delay, during which both levers are retracted. Rats used the reset lever adaptively on the PRR task, and their ratio reset decisions were sensitive to the cost of the reset delay. We derived an approach for computing the optimal bout length — the number of rewards to earn before pressing the reset lever that produces the greatest long-term rate of reward — and found that rats flexibly changed their behavior to approximate the optimal strategy. However, rats showed a systematic bias for bout lengths that exceeded the optimal length, an effect reminiscent of "overharvesting" in patch-leaving tasks. The PRR task thus represents a novel means of testing whether and how rats adapt their behavior to the cost-benefit structure of the environment in a way that connects deeply to the broader literature on associative learning and optimal foraging theory. ### Competing Interest Statement The authors have declared no competing interest. Whitehall Foundation, https://ror.org/00her7k05, Research Grant National Institutes of Health, 1R01MH137276, 2R01 DA047870 Brain and Behavioral Research Foundation, Young Investigator Award National Science Foundation, https://ror.org/021nxhr62, GRFP Howard Hughes Medical Institute, https://ror.org/006w34k90, Gilliam program University of California Office of the President
Dan Goodman

How do babies and blind people learn to localise sound without labelled data? We propose that innate mechanisms can provide coarse-grained error signals to boostrap learning.

New preprint from @yang_chu.

arxiv.org/abs/2001.10605

Thread below 👇

arXiv.orgLearning spatial hearing via innate mechanismsThe acoustic cues used by humans and other animals to localise sounds are subtle, and change during and after development. This means that we need to constantly relearn or recalibrate the auditory spatial map throughout our lifetimes. This is often thought of as a "supervised" learning process where a "teacher" (for example, a parent, or your visual system) tells you whether or not you guessed the location correctly, and you use this information to update your map. However, there is not always an obvious teacher (for example in babies or blind people). Using computational models, we showed that approximate feedback from a simple innate circuit, such as that can distinguish left from right (e.g. the auditory orienting response), is sufficient to learn an accurate full-range spatial auditory map. Moreover, using this mechanism in addition to supervised learning can more robustly maintain the adaptive neural representation. We find several possible neural mechanisms that could underlie this type of learning, and hypothesise that multiple mechanisms may be present and interact with each other. We conclude that when studying spatial hearing, we should not assume that the only source of learning is from the visual system or other supervisory signal. Further study of the proposed mechanisms could allow us to design better rehabilitation programmes to accelerate relearning/recalibration of spatial maps.
#neuroscience#computationalneuroscience#compneuro
TrendingLive feeds
About