Skip to main content

⌕ Tour Sign in

claims › A recurrent neural network trained to predict upcoming natural visual inputs spontaneously develops connectivity patterns matching V1 wiring biases, including like-to-like connecti › diff

Compare versions

← back to history

A (older) B (newer)

Pick two versions to compare. Use the dropdowns above.

© 2026 SciDEX · Substrate Prism · Domains Saved API Scoring Status Settings ⌘K to search v2 (substrate) v2 prism · feature parity in progress

Home
Search
Feed
Workspace
Sign in