Cortex and cerebellum: two papers on motor learning and representation change

Stable control under delayed feedback challenges both robotics and neuroscience. Classical cerebellar models explain delay compensation through forward prediction, but they do not account for the fast online corrections and rapid adaptation seen in biological systems.

Prediction plus feedback

Ana Carolina Filipe, Rui Ponte Costa, and Cláudia Soares propose a cerebellum-inspired control scheme. It combines multiplexed predictive representations with internal feedback. The encoder shares kinematic variables and task-relevant error signals, allowing precise movement correction despite delayed feedback. The authors show that adding feedback to the cerebellar loop alone significantly speeds up adaptation, cutting learning time by an order of magnitude. However, under delay, predictions based on a single signal are insufficient. Only multiplexing and feedback together provide a consistent mechanism for simultaneous online control and fast learning.

Changing representations without stimulation

The second paper addresses a different question. In neuropsychiatry, the goal is often not to read brain activity but to change it, such as weakening a negative affective charge or an overly consolidated memory. Marco Rothernel's team proposes a machine-learning framework of neural surrogate models. It suggests candidate representation changes and tests predicted perceptual effects using images of stimulus-evoked fMRI activity, without physical stimulation. The scheme combines fMRI decoding, deep generative modelling, and constrained control in latent space. Valence and memorability serve as working examples.

The data came from more than 36,000 image-fMRI observations from four deeply scanned participants in the Natural Scenes Dataset. Person-specific models reproduced the coarse generative structure of visual cortex, achieving 0.79 to 0.88 in a two-way identification task against a chance level of 0.5. Graded perturbations were reconstructed as images and assessed by automated scorers and human ratings from 7,200 trials performed by 18 participants. In the base model, valence shifted from -0.61 to +1.03 standard deviations, and memorability from -1.34 to +1.45 standard deviations. In human ratings, valence shifted in the predicted direction, with a mean slope of 0.038 standard deviations per unit of coefficient, a confidence interval of 0.003 to 0.074, and a positive slope in 16 of 18 participants. Perceived memorability did not change reliably.

The authors highlight the limitations. Agreement with the automated scorer was only suggestive for valence and weak for memorability. Extreme perturbations drifted from the original stimulus, so the intended change must be weighed against loss of fidelity. This is not a ready treatment, but a testable tool for designing and behaviourally testing candidate targets for future neuromodulation.