In recent years, most exciting inputs (MEIs) synthesized from encoding models of neuronal activity have become an established method for studying tuning properties of biological and artificial visual systems. However, as we move up the visual hierarchy, the complexity of neuronal computations increases. Consequently, it becomes more challenging to model neuronal activity, requiring more complex models. In this study, we introduce a novel readout architecture inspired by the mechanism of visual attention. This new architecture, which we call attention readout, together with a data-driven convolutional core outperforms previous task-driven models in predicting the activity of neurons in macaque area V4. However, as our predictive network becomes deeper and more complex, synthesizing MEIs via straightforward gradient ascent (GA) can struggle to produce qualitatively good results and overfit to idiosyncrasies of a more complex model, potentially decreasing the MEI's model-to-brain transferability. To solve this problem, we propose a diffusion-based method for generating MEIs via Energy Guidance (EGG). We show that for models of macaque V4, EGG generates single neuron MEIs that generalize better across varying model architectures than the state-of-the-art GA, while at the same time reducing computational costs by a factor of 4.7x, facilitating experimentally challenging closed-loop experiments. Furthermore, EGG diffusion can be used to generate other neurally exciting images, like most exciting naturalistic images that are on par with a selection of highly activating natural images, or image reconstructions that generalize better across architectures. Finally, EGG is simple to implement, requires no retraining of the diffusion model, and can easily be generalized to provide other characterizations of the visual system, such as invariances. Thus, EGG provides a general and flexible framework to study the coding properties of the visual system in the context of natural images.

The neural origin of the steady-state vergence eye movement error, called binocular fixation disparity, is not well understood. Further, there has been no study that quantitatively relates the dynamics of the vergence system to its steady-state behavior, a critical test for the understanding of any oculomotor system. We investigate whether fixation disparity can be related to the dynamics of opponent convergence and divergence neural pathways. Using binocular eye movement recordings, we first show that opponent vergence pathways exhibit asymmetric angle-dependent gains. We then present a neural model that combines physiological properties of disparity-tuned cells and vergence premotor cells with the asymmetric gain properties of the opponent pathways. Quantitative comparison of the model predictions with our experimental data suggests that fixation disparity can arise when asymmetric opponent vergence pathways are driven by a distributed disparity code.