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
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 changed | Where | Measured | Effect |
|---|---|---|---|
| Use XNOR_AIL instead of ReLU | small MLP · Synthetic parity | classification accuracy: 60% -> perfect classification accuracy | Improved |
| Use normalized XNOR operators rather than unnormalized XNOR operators (ablation) | transfer-learning MLPs · Image classification transfer tasks | test accuracy | Improved |
| Use OR_AIL rather than exact OR_IL (ablation) | transfer-learning MLPs · Image classification transfer tasks | test accuracy | Improved |
| Use XNOR_AIL alone in a deep ResNet | ResNet50 · CIFAR-10 and CIFAR-100 | classification performance | Hurt |
| Use OR_AIL rather than ReLU | ResNet50 · CIFAR-10 and CIFAR-100 | classification performance | Improved |
| Use a duplication ensemble rather than a partition ensemble (ablation) | AIL activation ensembles · Transfer learning | test accuracy | Improved |
| Combine OR, AND, and XNOR_AIL in an ensemble | 2-layer MLP · Bach Chorales | accuracy | No clear effect |
| Use XNOR-shaped activations at widths below 2e5 (ablation) | MLP · MNIST | classification performance | Hurt |
Closest papers by meaning
- Neuron Dependency Graphs: A Causal Abstraction of Neural Networks (ICML 2022 · similarity 0.86)
- BOLD: Boolean Logic Deep Learning (NeurIPS 2024 · similarity 0.85)
- Evaluating Neuron Explanations: A Unified Framework with Sanity Checks (ICML 2025 · similarity 0.84)
- Complex priors and flexible inference in recurrent circuits with dendritic nonlinearities (ICLR 2024 · similarity 0.84)
- Isotropic Activation Functions Enable Deindividuated Neurons and Adaptive Topologies (NeurIPS 2026 · similarity 0.84)
- Mind the Gap: Removing the Discretization Gap in Differentiable Logic Gate Networks (NeurIPS 2025 · similarity 0.84)
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.
