They report that eye-position gain fields are inaccurate immediately following a saccade, yet strikingly, saccadic behavior during that same interval see more remains accurate. From this, Xu et al. (2012) provocatively conclude that eye-position gain fields are not updated fast enough to be used by the brain to compute the location of targets for upcoming saccades. Gain fields underlie a prominent model for how spatial information is handled by the brain. According to this model, the oculomotor system combines retinal target information and eye-position information together in a distributed, population encoding of supraretinal

target location (Zipser and Andersen, 1988). The term “gain field” characterizes the way in which rate-coded postural signals (such as those carrying information about eye or hand position) interact with a receptive field or radial basis function (Poggio and Girosi, 1990). CH5424802 solubility dmso In particular, these rate-coded postural signals modulate the sensitivity or gain of an individual neuron’s response without otherwise changing (i.e., shifting, broadening, or sharpening) the neuron’s receptive field. For example, a neuron may be highly responsive when a visual stimulus is presented in its receptive field and the subject’s

gaze is to the right, yet respond only weakly when the same stimulus is presented in the receptive field and the subject’s gaze is to the left. The overall pattern of modulation of visual responses over a range of different eye positions constitutes the

neuron’s gain field. Eye-position gain fields were first observed in areas 7a and LIP of the parietal cortex (Andersen and Mountcastle, 1983).They have since been described in a wide range of cortical and subcortical areas including V1, V3A, V4, V6A, MT/MST, VIP, PMd, SEF, SC, and the LGN. The gain field model relies on population coding. Even though individual gain-modulated neurons receive the necessary inputs to represent target locations in supraretinal (e.g., head-centered) coordinates, this information is stored in a way that is ambiguous at the single-neuron level mafosfamide since many different combinations of eye position and retinal target location can give rise to the same neuronal response. The ambiguity is resolved by considering a population of neurons containing a broad distribution of gain fields and receptive fields. The representation of head-centered target information is thus implicit in the distributed population activity, rather than being explicitly represented by individual neurons with supraretinal receptive fields. An explicit representation by head-, body-, or world-centered neurons might appear to be a more efficient scheme than an implicit population encoding, since the explicit representation obviates the need for updating after each saccade, head, or body movement. However, behavioral and electrophysiological data reveal representations primarily based on eye-centered receptive fields (Baker et al.