Search

Humans explore the world by optimistically directing choices to less familiar options and by choosing more randomly when options are uncertain. Here, the authors show that these two exploration strategies rely on distinct uncertainty estimates represented in different parts of the prefrontal cortex.

Dopamine neurons encode reward prediction errors (RPE) that report the mismatch between expected reward and outcome for a given state. Here the authors report that when there is uncertainty about the current state, RPEs are calculated on the probabilistic representation of the current state or belief state.

The authors show how predictive representations are useful for maximizing future reward, particularly in spatial domains. They develop a predictive-map model of hippocampal place cells and entorhinal grid cells that captures a wide variety of effects from human and rodent literature.

How the human visual system leverages the rich structure in object motion for perception remains unclear. Here, Bill et al. propose a theory of how the brain could infer motion relations in real time and offer a unifying explanation for various perceptual phenomena.

The authors show that behavior and dopamine responses track prospective, but not retrospective, contingency. These results are explained by temporal difference (TD) learning models, indicating that TD errors are a measure of contingency.

Fan et al. show that trait somatic anxiety is associated with reduced tendency for exploration, including being less likely to choose an uncertain option.

Dopamine signals are implicated in not only reporting reward prediction errors but also various probabilistic computations. In this Opinion article, Gershman and Uchida propose that these different roles for dopamine can be placed within a common reinforcement learning framework.

McNamee et al. develop a theory of entorhinal–hippocampal processing. Distributed entorhinal input drives hippocampal activity between distinct statistical and dynamical regimes of activity, thereby unifying several empirical observations.

A long-standing idea in modern neuroscience is that the brain computes inferences about the outside world rather than passively observing its environment. The authors record from midbrain dopamine neurons during tasks with different reward contingencies and show that responses are consistent with a learning rule that harnesses hidden-state inference.

People are disproportionately more patient when evaluating larger rewards. Here, the authors show how this magnitude effect may reflect an adaptive response to uncertainty in mental representations of future value.

The authors show that in a creative setting people make choices partially based on how empowering their choices are.

A longitudinal study over 12 weeks used computational models on behavioural data from seven cognitive tasks while tracking participants’ mood, habits and activities to understand individual variability. The findings revealed that practice and emotional states significantly influenced various aspects of computational phenotypes, suggesting that apparent unreliability might actually uncover previously unnoticed patterns, supporting a dynamic perspective on cognitive diversity within individuals.

Previous work decoding linguistic meaning from imaging data has generally been limited to a small number of semantic categories. Here, authors show that a decoder trained on neuroimaging data of single concepts sampling the semantic space can robustly decode meanings of semantically diverse new sentences with topics not encountered during training.

Pavlovian and instrumentally driven actions often conflict when determining the best outcome. Here, the authors present an arbitration theory supported by human behavioral data where Pavlovian predictors drive action selection in an uncontrollable environment, while more flexible instrumental prediction dominates under conditions of high controllability.

Studying behaviour in a decision-making task with multiple features and changing reward functions, Tomov et al. find that a strategy that combines successor features with generalized policy iteration predicts behaviour best.

A Bayesian model incorporating representational splitting explains better memory performance in blocked compared to interleaved learning contexts.

The hypothesis that dopamine reports reward prediction errors has been both influential and controversial. This Perspective characterizes the present state of evidence, indicating where it succeeds and where it falls short. A complete account of dopamine will probably need to move beyond the reward prediction error hypothesis while retaining its core explanatory power.

This study reveals that dopamine is necessary for devaluing sensory memories of reward and thus plays a more complex role in reinforcement learning than traditionally considered.

To study cognition, researchers have traditionally used laboratory-based experiments, but games offer a valuable alternative: they are intuitive and enjoyable. In this Perspective, Schulz et al. discuss the advantages and drawbacks of games and give recommendations for researchers.

Visual working memory models combining deep neural network features with the Target Confusability Competition model capture human memory errors on GAN-generated scenes. Layers from the DNN architectures reproduce set-size effects and response bias curves in orientation and colour.