Get Started
Research questionWhat can finite threshold networks represent, and how do recurrence and invariance constrain that computation?The challenge is to relate network structure and finite-state dynamics to the logical functions and stimulus invariances neural systems can realize. Simplified response analyses can also miss selectivity when nonlinear neural responses have weak or vanishing first-order effects.
AI
Neural and Evolutionary Computing
Research Paper
Latest papersRecent research connected to this question, newest first.Neural Logic, Invariance, and the Retina---McCulloch and PittsApplies to the chapter’s historical and mathematical treatment of binary neural networks, threshold and veto inhibition, recurrent trajectories, group-related inputs, and spike-triggered analysis of frog-retina responses. Conclusions are limited by the stated idealizations and historical evidence.research paper · Sep 2, 2026
Related questions
When do explicitly organized recurrent interactions improve dynamical learning over generic reservoirs under matched state dimensions and controlled tuning?How do initialization correlation decay and feature-function Hermite rank determine residual networks’ scaling and continuous-depth limit?How do input dimension and network depth affect the tractability of verifying properties of ReLU networks?How can neural networks retain capacity while fitting within fixed parameter and memory budgets?
Home
Topics
Search
Library