W. Jeffrey Johnston

High-dimensional hippocampus representation facilitates learning of low-dimensional abstract representation and supports generalization

This paper shows that high-dimensional hippocampal representations aid in forming low-dimensional abstract representations and improve generalization.

The effects of task similarity during representation learning in brains and neural networks

This study shows that task similarity during learning can impair performance, with neural and network models showing reduced orthogonality in representations for similar tasks.

Impact of Training Regimes and Task Similarity on Learning and Cognitive Transfer

The paper investigates how task similarity and training regime affect learning and transfer, finding that blocked training improves performance on similar tasks, while interleaving hinders it, with an asymmetric effect based on task order.

When and why modular representations emerge

This study shows that modular representations in neural networks emerge for complex behavior when inputs are low-dimensional, without anatomical constraints, and supports rapid learning and generalization.

The neural basis of swap errors in working memory

This study identifies the neural mechanisms of swap errors in working memory by analyzing neural population recordings in monkeys performing multi-stimulus tasks.

Redundant representations are required to disambiguate simultaneously presented complex stimuli

The paper addresses the assignment problem in sensory integration by analyzing how redundant representations help disambiguate complex stimuli across modalities and brain regions.

Abstract representations emerge naturally in neural networks trained to perform multiple tasks

This paper shows that training neural networks on multiple tasks leads to the emergence of abstract representations, which support generalization and few-sample learning.

Primate superior colliculus is engaged in abstract higher-order cognition

This study identifies a novel role of the primate superior colliculus in abstract visual categorization, showing it encodes category information more strongly and quickly than the posterior parietal cortex.

The lateral intraparietal area preferentially supports stimulus selection in directed tasks compared to undirected free behavior

This study compares neural activity in the lateral intraparietal area (LIP) during directed and undirected tasks, finding that LIP encodes eye movement prior only in directed tasks.

Nonlinear mixed selectivity supports reliable neural computation

The paper shows nonlinear mixed selectivity enables reliable neural computation with lower metabolic energy than pure selectivity.