Hüseyin Boyacı and colleagues have published “A recurrent neuronal model for the effect of predictions on sensory processes” in Scientific Reports (October 2026).
The study proposes a new computational model of how expectations shape what we see. In the model, populations of neurons combine incoming sensory signals with feedback from higher brain areas. Fitted to behavioral data from an experiment with faces and houses, it reproduces how prior cues change perception, offering an alternative to the widely used predictive coding account.
Authors: Buse Merve Ürgen, Hilal İrem Baştürk, Hüseyin Boyacı