AI Paper Atlas

Research map / Architectures and efficiency / RNNs and normalization

Logical Activation Functions: Logit-space equivalents of Probabilistic Boolean Operators

Scott C Lowe, Robert Earle, Jason d'Eon, Thomas Trappenberg, Sageev Oore

NeurIPS 2022 · Architectures and efficiency › RNNs and normalization · approach: Mechanistic circuit interpretability · 0 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 XNOR_AIL instead of ReLUsmall MLP · Synthetic parityclassification accuracy: 60% -> perfect classification accuracyImproved
Use normalized XNOR operators rather than unnormalized XNOR operators (ablation)transfer-learning MLPs · Image classification transfer taskstest accuracyImproved
Use OR_AIL rather than exact OR_IL (ablation)transfer-learning MLPs · Image classification transfer taskstest accuracyImproved
Use XNOR_AIL alone in a deep ResNetResNet50 · CIFAR-10 and CIFAR-100classification performanceHurt
Use OR_AIL rather than ReLUResNet50 · CIFAR-10 and CIFAR-100classification performanceImproved
Use a duplication ensemble rather than a partition ensemble (ablation)AIL activation ensembles · Transfer learningtest accuracyImproved
Combine OR, AND, and XNOR_AIL in an ensemble2-layer MLP · Bach ChoralesaccuracyNo clear effect
Use XNOR-shaped activations at widths below 2e5 (ablation)MLP · MNISTclassification performanceHurt

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