Alon Boaz Baram

A mechanistic theory of planning in prefrontal cortex

This paper proposes a mechanistic theory of planning in prefrontal cortex using spacetime attractors, showing that recurrent neural networks trained on planning tasks recapitulate this structure.

An abstract relational map emerges in the human medial prefrontal cortex with consolidation

The paper investigates the emergence of abstract relational representations in the human medial prefrontal cortex over time using fMRI and graph-learning paradigms.

Flexible neural representations of abstract structural knowledge in the human Entorhinal Cortex

The paper demonstrates that the human entorhinal cortex represents abstract structural knowledge in non-spatial tasks with hexagonal grid structure, but not in tasks with clustered structure.

A hierarchical coordinate system for sequence memory in human entorhinal cortex.

The paper shows that a hierarchical coordinate system in the human entorhinal cortex supports abstract sequence memories, mirroring spatial memory organization.

Generative replay for compositional visual understanding in the prefrontal-hippocampal circuit

The paper shows that visual understanding involves relational and compositional representations in the prefrontal-hippocampal circuit, with replay sequences forming hypotheses about scene configurations.

Entorhinal and ventromedial prefrontal cortices abstract and generalise the structure of reinforcement learning problems

The paper shows that entorhinal and ventromedial prefrontal cortex representations generalize reinforcement learning problem structures across different tasks.

Intuitive planning: global navigation through cognitive maps based on grid-like codes

The paper proposes that grid-like codes can encode global navigation information through cognitive maps, enabling intuitive planning without extensive lookahead.

What is a cognitive map? Organising knowledge for flexible behaviour

The paper reviews evidence and theories suggesting cognitive maps organize knowledge across domains, supporting flexible behavior and abstract learning.