AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Deciding How to Decide: Dynamic Routing in Artificial Neural Networks

Mason McGill, Pietro Perona

ICML 2017 · Architectures and efficiency › RNNs and normalization · approach: Efficient network compression and training · 55 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 dynamic routing instead of static routing at a fixed computational budgetImage-classification networks · Hybrid MNIST/CIFAR-10accuracyImproved
Use actor learning instead of critic learningDynamically-routed networks · Hybrid MNIST/CIFAR-10accuracyImproved
Use optimistic critic learning instead of pragmatic critic learning (ablation)Dynamically-routed networks · Hybrid MNIST/CIFAR-10accuracyHurt
Replace cross-entropy error with classification error in the pragmatic critic utility-regression term (ablation)Pragmatic critic networks · Hybrid MNIST/CIFAR-10accuracyHurt
Add regularization on routing decisions (ablation)Dynamically-routed networks · Hybrid MNIST/CIFAR-10peak accuracyHurt
Add a nonzero decisiveness cost to encourage exploration (ablation)Actor networks · Hybrid MNIST/CIFAR-10training effectivenessImproved
Apply throughput-adjusted learning rates (ablation)Actor and critic networks · Hybrid MNIST/CIFAR-10, computational-resource-abundant high-accuracy contextsperformanceImproved
Use junctions with both 2 and 3 branches rather than only 2-way junctions (ablation)Dynamically-routed networks · Hybrid MNIST/CIFAR-10 under tight computational constraintsaccuracyImproved
Use higher-degree branching (ablation)Dynamically-routed networks · Hybrid MNIST/CIFAR-10 without computational constraintsoverfittingHurt
Train one actor network conditioned on computation cost instead of separate networks for fixed costs (ablation)Actor network · Hybrid MNIST/CIFAR-10 across varying computational constraintsmemory consumption and training time: 8-fold reductionImproved
Use dynamic routing instead of static routingActor networks · CIFAR-2peak accuracyImproved
Use dynamic routing instead of static routingActor networks · CIFAR-10accuracyMixed

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