Human speech carries information beyond the words themselves: pitch, loudness, duration, and pauses–jointly referred to as ‘prosody’–emphasize critical words, help group words into phrases, and convey emotional and other socially-relevant information. Using a novel fMRI paradigm, we identify a set of prosody-responsive brain areas and then richly characterize them across 8 experiments (25 experimental conditions; n=51 participants). These areas–located on the lateral temporal surface bilaterally and in the frontal lobe–respond to the presence of prosody in both meaningful and meaningless speech, and are distinct from nearby temporal pitch- and speech-perception areas and from frontal areas sensitive to general cognitive and attentional demands. They are also dissociable from–but show partial overlap with–the language areas and with the area that processes dynamic facial expressions. Thus, prosodic information is processed by a distinct set of brain areas, which may help integrate linguistic meanings with non-verbal communicative signals.
Magnification-Aware Distillation (MAD): A Self-Supervised Framework for Unified Representation Learning in Gigapixel Whole-Slide Images
arXiv:2512.14796v1 Announce Type: cross Abstract: Whole-slide images (WSIs) contain tissue information distributed across multiple magnification levels, yet most self-supervised methods treat these scales as independent

