How does the similarity structure of memory influence credit assignment in reinforcement learning? Memory spaces vary in how integrated versus separable their constituent dimensions are, and how clustered versus distributed items are across dimensions. Greater integration may cause people to overattribute value to multiple dimensions, potentially leading to generalization errors. Greater clustering may bias people to attribute value to discrete category-centers (category-based learning) rather than map value continuously across the space (function learning). In this study, subjects complete a value-learning task in which stimuli are sampled from low-dimensional perceptual spaces, and reward is mapped to one dimension. Each space is intended to engender a different degree of integration and clustering in memory, such that effects of memory organization on learning can be probed. Additionally, we investigate how credit assignment on each of these artificial perceptual spaces differs to credit assignment on more complex spaces that define real-world semantic concepts.

Past work has made conflicting proposals about the mechanisms underlying instructed reinforcement learning (RL)specifically,that prefrontal cortex, representing instruction, either biases, attenuates, or overrides learning signals in the brain. Weleverage the sensitivity of pattern-similarity analysis of fMRI data to distinguish between the qualitative features of theseaccounts. Participants learn the value of six novel stimuli after receiving false information that one is of high value. Wetrack markers of value learning in visual cortex during a value-independent perceptual judgement task presented betweenintervals of RL. We predict that with learning, the correlation between activation patterns for similarly valued stimuli willincrease. To characterize influences on learning, we examine how the rate at and direction in which these patterns changein similarity will be influenced by explicit instruction about stimulus value. This work will help us identify the principlecognitive and neural mechanisms underlying instructed RL.