During the first period, we (1) identified a core mechanism that stabilizes vision during movement (Vega-Zuniga et al., 2025). We uncovered a hub-and-spoke network centered in the ventral lateral geniculate nucleus (vLGN) that integrates motor signals to distinguish self-generated from external sensory events (Aim 2a,b).
In the second period, we (2) extended our work on adaptation versus invariance in early vision (Aim 1a,b), showing that luminance shifts spike timing by up to 25 ms, while relative spike timing remains stable—providing an invariant code for perception under natural conditions (Gupta, Patel, et al. in preparation). In parallel, we (3) discovered intrinsic, stimulus-independent population dynamics in the superior colliculus (SC) that track learning, attention, and performance, and interact with sensory input on single trials to shape visual encoding (Schmidt et al., in submission; Aim 1a,b) (Schmidt, et al. in submission). Focusing on neuromodulation, we (4) showed that noradrenaline from the locus coeruleus shifts the excitatory–inhibitory balance in SC circuits, likely gating specific visuomotor transformations (Aim 2c, 3a,b). We further (5) investigated serotonergic dysregulation in ASD, revealing sensory deficits, though linking these to behavior remains ongoing (Aim 2c).
Currently, we are (6) initiating studies of dopaminergic signaling in the SC (Aim 3a,b) and (7) developing theoretical frameworks of invariance to understand how stable decoding is maintained across changing neuromodulatory states (Aim 1a,b).
Finally, (8) as part of a side project, we contributed to retinal circuit studies with Yi-Rong Peng (UCLA), shedding light on the connectivity logic that gives rise to computation (Cavallini, Quiroli, et al., submitted), with insights and approaches developed during the ERC supporting this work.
