Skip to content
DraconiticResearch
Search Sources API
203,203 studies31 sourcesUpdated 4 min ago
Peer-reviewedCC BY-NC-NDNature Communications

A single computational objective can produce specialization of streams in visual cortex

  • Dawn Finzi1,2
  • Eshed Margalit3
  • Kendrick Kay4
  • Daniel L. K. Yamins5,6,7
  • Kalanit Grill-spector A6,7
7 affiliations
  1. 1Department of Computer Science, Stanford University, Stanford, CA, USA. [email protected]
  2. 2Department of Psychology, Stanford University, Stanford, CA, USA. [email protected]
  3. 3Neurosciences Graduate Program, Stanford University, Stanford, CA, USA
  4. 4Center for Magnetic Resonance Research, Department of Radiology, University of Minnesota, Minneapolis, MN, USA
  5. 5Department of Computer Science, Stanford University, Stanford, CA, USA
  6. 6Department of Psychology, Stanford University, Stanford, CA, USA
  7. 7Wu Tsai Neurosciences Institute, Stanford University, Stanford, CA, USA
Open at sourcePDFDOICite

Abstract

Human visual cortex is organized into dorsal, lateral, and ventral streams. A long-standing hypothesis is that the functional organization into streams emerged to support distinct visual behaviors. Here, we compare neural network-based computational models against a massive fMRI dataset to investigate why visual streams emerge. We find that a self-supervised Topographic Deep Artificial Neural Network (TDANN), which encourages nearby units to respond similarly, better captures brain responses, as well as spatial segregation and functional differentiation across streams, than DANN models trained for stream-specific visual behaviors. These findings challenge the prevailing view that streams evolved to separately support different behaviors and suggest instead the possibility that functional organization can arise from a single principle: learning generally useful visual representations subject to local spatial constraints. Visual streams are often explained as pathways optimized for different behaviors. Here, the authors show that brain-like stream organization can instead arise from a single self-supervised objective under local spatial constraints.

Related

  1. ReGraph: A Computational Account of Emergent Generalization in the "what" and "where" Dual Visual Streams
    Hyewon Kang, Jungmin Lee, Ilgyu Lee +1
    arXivPreprintarXiv licenceFull textNeurosciencearXiv
  2. SPECHTacle: An Efficient Detection and Visualization of Cell Specific Regulation in Single Cell-Based Experiments Using a Novel Combinatorial Computational Tool
    K. M. Specht, H. C. Schwierling, M. J. Robson
    bioRxivPreprintCC BY-NDFull textBioinformatics & computational biologybioRxiv
  3. Beyond a single threshold: calibration of CADD v1.7 and AlphaGenome for ACMG/AMP computational evidence across variant classes, model versions, and reference data
    J. P. Miranda, L. Carvajal, I. Aldana +2
    bioRxivPreprintCC BYFull textGenetics & genomicsbioRxiv
  4. Varied forms of mutant tRNA mistranslation produce distinct phenotypes across multiple model organisms
    B. S. Charles, A. J. Moehring
    bioRxivPreprintAll rights reservedFull textGenetics & genomicsbioRxiv
  5. Transcriptional regulatory programs underlying cellular and regional specialization in the late-gestation porcine brain
    Jiacheng Wei, Ping Liu, Siyu Yang
    Communications BiologyPeer-reviewedCC BY-NC-ND 4.0Full textNeuroscienceCrossref

Cite

BibTeXRIS
BibTeX
@article{finzi2026single,
  title = {{A single computational objective can produce specialization of streams in visual cortex}},
  author = {Dawn Finzi and Eshed Margalit and Kendrick Kay and Daniel L. K. Yamins and Kalanit Grill-spector A},
  journal = {Nature Communications},
  year = {2026},
  month = aug,
  doi = {10.1038/s41467-026-76098-y},
  url = {https://doi.org/10.1038/s41467-026-76098-y},
  note = {CC BY-NC-ND},
}
RIS
TY  - JOUR
TI  - A single computational objective can produce specialization of streams in visual cortex
AU  - Dawn Finzi
AU  - Eshed Margalit
AU  - Kendrick Kay
AU  - Daniel L. K. Yamins
AU  - Kalanit Grill-spector A
PY  - 2026
DA  - 2026/08/13
JO  - Nature Communications
DO  - 10.1038/s41467-026-76098-y
UR  - https://doi.org/10.1038/s41467-026-76098-y
AB  - Human visual cortex is organized into dorsal, lateral, and ventral streams. A long-standing hypothesis is that the functional organization into streams emerged to support distinct visual behaviors. Here, we compare neural network-based computational models against a massive fMRI dataset to investigate why visual streams emerge. We find that a self-supervised Topographic Deep Artificial Neural Network (TDANN), which encourages nearby units to respond similarly, better captures brain responses, as well as spatial segregation and functional differentiation across streams, than DANN models trained for stream-specific visual behaviors. These findings challenge the prevailing view that streams evolved to separately support different behaviors and suggest instead the possibility that functional organization can arise from a single principle: learning generally useful visual representations subject to local spatial constraints. Visual streams are often explained as pathways optimized for different behaviors. Here, the authors show that brain-like stream organization can instead arise from a single self-supervised objective under local spatial constraints.
SN  - 2041-1723
ER  -