Keynote speakers:

Li Zhaoping, University of Tuebingen; Max Planck Institute, Germany

What does the primary visual cortex (V1) do?

Abstract: Decades after Hubel and Wiesel's discovery of V1' neural properties, we were asking questions such as "Do we know what the early visual system does?" (Carandini et al 2005) and "What is the other 85 percent of V1 doing?" (Olshausen and Field 2006), largely because the traditional understanding of how V1 neurons respond to visual inputs accounts for only a small fraction of V1's neural activities. Instead of focusing on how a V1 neuron responds to sensory inputs, we ask what V1 does for visual behavior. I will show converging behavioral and neural evidence for the following two functional roles for V1 in primates (and future studies are likely to reveal and elucidate additional roles). First, according to the V1 Saliency Hypothesis (V1SH), V1 creates a bottom-up saliency map of the visual field by its neural responses (Li 2002). This saliency map guides shifts of gaze (attention) via superior colliculus which receives monosynaptic V1 projections. Second, through feedforward and feedback projections between visual cortical areas, and according to a dichotomy between the central and peripheral visual fields, V1 works with higher visual areas to achieve visual recognition in light of the attentional bottleneck that starts at V1's output to higher visual areas. Specifically, the feedforward signals from V1 to higher visual areas are impoverished by the bottleneck, making visual recognition vulnerable to ambiguities and illusions; to overcome, top-down feedback queries for additional information from V1 (and likely other early visual areas) to aid ongoing recognition, and, due to limited brain resources, this top-down query is restricted mainly to the central visual field according to the Central-peripheral dichotomy (CPD) theory (Zhaoping 2019, 2025). This V1SH-Bottleneck-CPD framework gives non-trivial predictions. Confirmed predictions include gaze shifts evoked by invisible eye-of-origin visual features, V1 neural responses predictive of subsequent saccades, and newly predicted visual illusions that are typically visible only in the peripheral visual field and become visible in the central visual field when the feedback query is compromised.

References:

Carandini, et al, 2005. Do we know what the early visual system does?. Journal of Neuroscience, 25(46), pp.10577-10597. Olshausen, B.A. and Field, D.J., 2006. What is the other 85 percent of V1 doing. In ``Problems of Systems Neuroscience", L. van Hemmen, & T. Sejnowski (Eds.), 23, pp.182-211. Li, Z., 2002. A saliency map in primary visual cortex. Trends in cognitive sciences, 6(1), pp.9-16. Zhaoping, L., 2019. A new framework for understanding vision from the perspective of the primary visual cortex. Current opinion in neurobiology, 58, pp.1-10. Zhaoping, L., 2025. Testing the top-down feedback in the central visual field using the reversed depth illusion. iScience.

Sacha van Albada, Jülich Research Centre, Germany

Large-scale spiking models of the cerebral cortex: from anatomy to dynamics

Abstract: Simulations form a third pillar of science next to experiments and theory. In neuroscience, simulations can provide insight into relationships between brain anatomy and the dynamics of neural circuits. The large numbers of neurons and astronomical numbers of synapses of mammalian neural circuits have historically prevented full-scale simulations of these circuits resolving the individual neurons and synapses. Modern simulation technology and supercomputers now make such full-scale simulations possible. Leveraging these technological advances and integrating extensive anatomical data, we have developed spiking network models of the cerebral cortex featuring the full density of neurons and synapses in each local circuit. The models reveal how structural network properties influence ongoing cortical activity, from the level of spikes to that of multiple areas. By making the model codes publicly available, we hope that the work forms a basis for ever more refined and predictive brain models involving the cerebral cortex.