Neuron to Graph: Interpreting Language Model Neurons at Scale [PDF]
Advances in Large Language Models (LLMs) have led to remarkable capabilities, yet their inner mechanisms remain largely unknown. To understand these models, we need to unravel the functions of individual neurons and their contribution to the network. This paper introduces a novel automated approach designed to scale interpretability techniques across a
arxiv
The Tumor Suppressor HHEX Inhibits Axon Growth when Prematurely Expressed in Developing Central Nervous System Neurons [PDF]
Neurons in the embryonic and peripheral nervoussystem respond to injury by activating transcriptional programs supportive of axon growth, ultimately resulting in functional recovery.
Blackmore, Murray G.+9 more
core +2 more sources
Hydrogen sulfide intervention in focal cerebral ischemia/reperfusion injury in rats
The present study aimed to explore the mechanism underlying the protective effects of hydrogen sulfide against neuronal damage caused by cerebral ischemia/reperfusion.
Xin-juan Li+6 more
doaj +1 more source
Single Biological Neurons as Temporally Precise Spatio-Temporal Pattern Recognizers [PDF]
This PhD thesis is focused on the central idea that single neurons in the brain should be regarded as temporally precise and highly complex spatio-temporal pattern recognizers. This is opposed to the prevalent view of biological neurons as simple and mainly spatial pattern recognizers by most neuroscientists today.
arxiv
Death and rebirth of neural activity in sparse inhibitory networks [PDF]
In this paper, we clarify the mechanisms underlying a general phenomenon present in pulse-coupled heterogeneous inhibitory networks: inhibition can induce not only suppression of the neural activity, as expected, but it can also promote neural ...
Angulo-Garcia, David+3 more
core +4 more sources
Should Under-parameterized Student Networks Copy or Average Teacher Weights? [PDF]
Any continuous function $f^*$ can be approximated arbitrarily well by a neural network with sufficiently many neurons $k$. We consider the case when $f^*$ itself is a neural network with one hidden layer and $k$ neurons. Approximating $f^*$ with a neural network with $n< k$ neurons can thus be seen as fitting an under-parameterized "student" network ...
arxiv
Differential regulation of the Drosophila sleep homeostat by circadian and arousal inputs. [PDF]
One output arm of the sleep homeostat in Drosophila appears to be a group of neurons with projections to the dorsal fan-shaped body (dFB neurons) of the central complex in the brain.
Elsheikh, Ahmed+7 more
core +1 more source
nNOS(+) striatal neurons, a subpopulation spared in Huntington's Disease, possess functional NMDA receptors but fail to generate mitochondrial ROS in response to an excitotoxic challenge. [PDF]
Huntington's disease (HD) is a neurodegenerative condition characterized by severe neuronal loss in the cortex and striatum that leads to motor and behavioral deficits.
Canzoniero, Lorella MT+5 more
core +2 more sources
We describe a novel set of Epac‐based FRET‐FLIM biosensors with improved fully cytosolic distribution, achieved without compromising the state‐of‐the‐art performance of our original designs, for detecting cAMP dynamics in real time in live cells with high precision and reliability.
Giulia Zanetti+2 more
wiley +1 more source
Well‐being on supportive techniques in amyotrophic lateral sclerosis: from neurologists' perspective
Abstract Objective To investigate intercultural neurologists' perception of well‐being in patients with amyotrophic lateral sclerosis (ALS) using gastrostomy (PEG), non‐invasive, and/or invasive ventilation (NIV/IV) and to analyse the determinants and impact on the management of the above medical interventions (MIs).
Krzysztof Barć+7 more
wiley +1 more source