AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Information dynamics and Memory in Neural Networks through Fisher Information Diffusion

Haodong Qin, Tatyana Sharpee

ICML 2026 · Architectures and efficiency › RNNs and normalization · approach: Brain-inspired recurrent architectures

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use Fisher-information-optimal block-structured initialization instead of Xavier, Kaiming, orthogonal, unitarysimple tanh RNN (100 neurons) · Copy task (Tdelay=50)convergence speed and accuracyImproved
use Fisher-optimal initialization with fixed recurrent weights, readout onlysimple tanh RNN · Sequential MNISTconvergence and accuracyImproved
use Fisher-optimal initialization with fixed recurrent weightssimple tanh RNN · CIFAR-10 and IndoorCVPR 09 sequential classificationconvergence and accuracyImproved
apply Fisher-optimal initialization to LSTM/GRU recurrent componentsLSTM/GRU gates · sequential memory taskstraining performanceImproved
operate at criticality (spectral radius 1) alone (ablation)two-population RNN · average Fisher information over 100 stepsretained Fisher informationMixed
add sparse feedback links to feedforward chain (ablation)chain RNN with 2-8 subpopulations · optimized Fisher informationFisher informationImproved
add indiscriminate feedback (ablation)chain RNN · optimized Fisher informationFisher informationHurt
increase number of subpopulations at fixed neuron count (ablation)optimized chain RNN · total Fisher informationFisher informationImproved
increase network size Nanalytic Fisher diffusion vs simulation · Fisher information trajectoriesmean-squared errorImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy