32 fMRI-scanned humans and 8 frontier open weight LLMs play ARC-AGI like games with no rules given. The reasoning models match the human learning trajectories and their hidden states predict human bra...

A wealth of evidence supports the view that neural manifolds are real and useful, even if they may not solve the age-old mind-body problem.

Inside Neuroscience · Episode

A series of NeuroAI themed talks organised by the UCL NeuroAI community. Talks will continue on a monthly basis.

Action embeddings for RL - model of motor adaptation and generalization - jessegeerts/action-embedding

Transformer-based models have demonstrated remarkable reasoning abilities, but the mechanisms underlying relational reasoning remain poorly understood. We investigate how transformers perform \textit{...

Aligning foundation models with human judgments enables them to more accurately approximate human behaviour and uncertainty across various levels of visual abstraction, while additionally improving th...

PhD position or a Doctoral Candidate in computational neuroscience. (m/f/d) | Karriereportal Max-Planck-Institut für Kognitions- und Neurowissenschaften

Humans and animals show remarkable learning efficiency, adapting to new environments with minimal experience. This capability is not well captured by standard reinforcement learning algorithms that re...

Transformer-based models have demonstrated remarkable reasoning abilities, but the mechanisms underlying relational reasoning in different learning regimes remain poorly understood. In this work, we i...

Movement-sensing neurons in the striatum influence a mouse’s choice of action by favoring routine behaviors.

Dopaminergic action prediction error signals are used by mice as a value-free teaching signal to reinforce stable sound–action associations in the tail of the striatum.