paperKB
coga / coga-kb
Help
Sign in

Chunk #19 — DISCUSSION

Source
States versus rewards: dissociable neural prediction error signals underlying model-based and model-free reinforcement learning.
Embedded
yes

Text

The finding that BOLD activity in pIPS correlates with an SPE may be interpreted in the context of previous neurophysiological studies into the activity of neurons in the lateral intraparietal area (LIP) during saccadic decision-making. Putative pyramidal cells report expectations about as-yet unknown characteristics about the state of the world (Gold and Shadlen, 2002), possibly coded in terms of the expected values of saccades (Platt and Glimcher, 1999; Sugrue et al., 2004). Other subregions of posterior parietal cortex (PPC) appear to be specialized for different movement modalities (Cui and Andersen, 2007), though less is known about their behavior with respect to decision variables. Under the view that the fMRI BOLD signal reflects in part the input into an area and intrinsic computations within it (Logothetis et al., 2001), it is straightforward to envision the SPE input that we recorded as being necessary for learning the structure of the environment necessary to support these predictions. The finding that SPE signals are present in pIPS while subjects are learning state transitions even in the complete absence of reward (session 1 of our