Back

Multifaceted brain representation of numerosity across the senses and presentation formats

Yang, Y.; Fornaciai, M.; Togoli, I.; Shahzad, I.; Gau, R.; Van Audenhaege, A.; Cerpelloni, F.; Collignon, O.

2025-06-16 neuroscience
10.1101/2025.05.29.653470 bioRxiv
Show abstract

Humans can extract numerosity from different senses and a variety of context. How and where the brain abstract numerical information from low-level sensory inputs remains debated. Using multivariate pattern decoding and representational similarity analysis applied to fMRI data, we comprehensively investigate how the brain represents numerical information (range 2-5) across different modalities (auditory, visual) and presentation formats (sequential, simultaneous; symbolic, non-symbolic). We identify a set of brain regions along the dorsal pathway-from early visual cortex to the intraparietal and frontal regions-that encode specific non-symbolic numerical information across formats and modalities. The numerical distance effect, a hallmark of magnitude encoding, was observed in most of these regions. We found aligned representation of numerical information across visual and auditory modalities in intraparietal and frontal regions, but only when they shared a sequential presentation format. Maintaining a distinction between spatial and temporal numerical representations may thus be a fundamental aspect of numerical processing. RSA further revealed a posterior-to-anterior gradient in the intraparietal sulcus (IPS) showing that the dominant factors influencing distributed numerical representations shifted from sensory modality in the posterior parietal regions to presentation format in the anterior parietal areas. Our study reveals a multifaceted brain representation of numerosity across the senses and presentation formats.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.