The Synaptic Report

Global Intelligence Database

Results for "Neuroscience"

Front Page Only
FOUND 60 MATCHES
#21Computer Science & AIFront Page10 November 2025

New 3D ConvNeXt Framework Boosts Whole-Brain fMRI Decoding with Enhanced Interpretability

A novel 3D ConvNeXt framework significantly improves whole-brain task fMRI decoding by integrating layer-global response normalization and efficient residual connections. This advancement offers robust, efficient, and interpretable insights into brain activity, paving the way for better cognitive neuroscience research and clinical applications like early diagnosis.

By Lim, Kim

#22Neuroscience9 November 2025

Brain-Inspired Networks Learn to Identify and Avoid Obstacles, Offering New Robotics and Disease Insights

Researchers developed a neuroscience-inspired neural network that uses synaptic synchronization and feedback inhibition to achieve robust pattern separation, a crucial cognitive function. This model not only provides insights into brain function and disorders like autism but also successfully guided a simulated robot to identify a trained stimulus as an obstacle and avoid it in its environment.

By Faghihi, Moustafa, Neymotin

#23NeuroscienceFront Page12 November 2025

How the Brain Builds a Timeline for Your Memories

Neuroscientists are synthesising evidence to understand how our brains track time for memories, a concept known as episodic time. They are exploring how specialised 'time cells' pinpoint specific moments, while other cells measure elapsed durations, creating a neural timeline for our experiences.

By Kwok, Lapate, Sakon, van Wassenhove, Xue, Zheng

#24NeuroscienceFront Page14 November 2025

Visualising the 3D Architecture of a Memory Circuit

Researchers have established a new technology to view the complex internal structure of memory-holding cells, known as engram cells. By combining advanced cryogenic microscopy techniques, they successfully mapped the 3D organisation of macromolecules within the synapses of a mouse brain.

By Lovatt, O’Sullivan, Ortega-de San Luis, Ryan, Frank

#25NeuroscienceFront Page13 November 2025

New AI Tool Teaches Digital Neurons to Behave Like Real Ones

Neuroscientists have a powerful new tool called JAXLEY for creating highly detailed models of brain cells. By harnessing modern AI techniques and GPU acceleration, this framework can efficiently teach digital neurons to mimic real biological behaviour or even perform complex computational tasks.

By Deistler, Kadhim, Pals, Beck, Huang, Gloeckler, Lappalainen, Schröder, Berens, Gonçalves, Macke

Page 3 of 6