Variation in few-shot classification error across predicted human visual cortex.

<p>All results are based on the outputs of our encoding model. <b>(A)</b> Average 5-shot error (1 - fraction of correctly classified exemplars; y-axis) as a function of WordNet path distance (x-axis; see inset) between the concept pairs in all ROIs (colored curves) considered here....

Full description

Saved in:

Bibliographic Details
Main Author:	Ghislain St-Yves (22238830) (author)
Other Authors:	Kendrick Kay (7901603) (author), Thomas Naselaris (103861) (author)
Published:	2025
Subjects:	Neuroscience Biotechnology Science Policy Biological Sciences not elsewhere classified Information Systems not elsewhere classified large fmri dataset g ., objects deep neural networks support accurate classification shot classification performance single geometric property simple &# 8220 shot &# 8221 brain area depends brain activity patterns estimated geometric properties accurate linear classification mid dnn layers concept manifolds differs geometric properties linear classification dnn layers &# 8217 concept manifolds linear classifiers visual cortex vice versa scenes ). recent theory negatively correlated even though encode images effective number directly determine
Tags:	Add Tag No Tags, Be the first to tag this record!

Be the first to leave a comment!

Variation in few-shot classification error across predicted human visual cortex.

Similar Items