AI Paper Atlas

Research map / Continual learning / Online and replay-based

Operative dimensions in unconstrained connectivity of recurrent neural networks

Renate Barbara Krause, Matthew Cook, Sepp Kollmorgen, Valerio Mante, Giacomo Indiveri

NeurIPS 2022 · Continual learning › Online and replay-based · approach: Gradient projection and plasticity preservation · 2 citations (OpenAlex)

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 operative dimensions instead of high-variance weight PCs to build reduced-rank WVanilla RNN, context-dependent integration · Context-dependent integrationdimensions needed for original performance: 88 -> 15Improved
Use operative dimensions instead of high-variance weight PCs to build reduced-rank WVanilla RNN, sine wave generation · Sine wave generationdimensions needed for original performance: 86 -> 29Improved
Define operative dimensions from random sampling locations rather than along condition-average trajectories (ablation)Operative dimensions · Context-dependent integration and sine wave generationnetwork output cost at fixed rankHurt
Use too few sampling locations along trajectories (ablation)Operative dimensions · Context-dependent integration and sine wave generationreproduction of network output and trajectoriesHurt
Add first few operative dimensions to weight matrixOperative-dimension reduced-rank RNN · Context-dependent integration and sine wave generationnetwork output costHurt
Increase number of input and output signals from 1 to 5Sine wave generator RNN · Multi-sine wave generationoperative dimensions neededMixed
Use operative dimensions from context 1 to run context 2 (and vice versa)Context-dependent integrator · Context-dependent integrationnetwork output costHurt
Use operative dimensions from one choice to run the other choiceContext-dependent integrator · Context-dependent integrationnetwork output costNo clear effect
Increase rank k of high-variance reduced WVanilla RNN · Context-dependent integrationnetwork output cost: k=60 to k=61Mixed

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