Untrained random CNNs align more closely with human V1 fMRI representations than backpropagation-trained models, while learning rules show differences mainly in intermediate areas.
P., Cownden, D., Tweed, D
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
fields
cs.LG 2years
2026 2verdicts
UNVERDICTED 2representative citing papers
Cross-species RSA finds conserved early visual cortex alignment across learning rules but divergent IT rankings, with model capacity and training data limiting higher-area matches more than the specific rule.
citing papers explorer
-
Untrained CNNs Match Backpropagation at V1: A Systematic RSA Comparison of Four Learning Rules Against Human fMRI
Untrained random CNNs align more closely with human V1 fMRI representations than backpropagation-trained models, while learning rules show differences mainly in intermediate areas.
-
Cross-Species RSA Reveals Conserved Early Visual Alignment but Divergent Higher-Area Rankings Across Human fMRI and Macaque Electrophysiology
Cross-species RSA finds conserved early visual cortex alignment across learning rules but divergent IT rankings, with model capacity and training data limiting higher-area matches more than the specific rule.