Research map / Learning theory and algorithms / Combinatorial and search optimization
Learning Logistic Circuits
Yitao Liang, Guy Van den Broeck
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 real-valued pixel inputs rather than binarizing pixels (ablation) | Logistic Circuit · MNIST test | accuracy (%): 97.4 -> 99.4 | Improved |
| use real-valued pixel inputs rather than binarizing pixels (ablation) | Logistic Circuit · Fashion test | accuracy (%): 87.6 -> 91.3 | Improved |
| use logistic-circuit structure learning instead of a 5-layer MLP | real-valued Logistic Circuit · MNIST, 2% training data | accuracy (%): 94.3 -> 96.1 | Improved |
| use logistic-circuit structure learning instead of a 5-layer MLP | real-valued Logistic Circuit · Fashion, 2% training data | accuracy (%): 80.9 -> 86.0 | Improved |
| use logistic-circuit structure learning instead of a CNN with 3 convolutional layers | real-valued Logistic Circuit · MNIST, 2% training data | accuracy (%): 95.3 -> 96.1 | Improved |
| use logistic-circuit structure learning instead of a CNN with 3 convolutional layers | real-valued Logistic Circuit · Fashion, 2% training data | accuracy (%): 83.8 -> 86.0 | Improved |
| use logistic-circuit structure learning instead of a 5-layer MLP | real-valued Logistic Circuit · MNIST test | accuracy (%): 99.3 -> 99.4 | No clear effect |
| use logistic-circuit structure learning instead of a 5-layer MLP | real-valued Logistic Circuit · Fashion test | accuracy (%): 89.8 -> 91.3 | Improved |
Closest papers by meaning
- A Unified Framework for Human-Allied Learning of Probabilistic Circuits (AAAI 2025 · similarity 0.83)
- On the Expressive Power of Tree-Structured Probabilistic Circuits (NeurIPS 2024 · similarity 0.82)
- The Computational Complexity of Circuit Discovery for Inner Interpretability (ICLR 2025 · similarity 0.81)
- Probabilistic Neural Circuits (AAAI 2024 · similarity 0.81)
- Neural Network Approximators for Marginal MAP in Probabilistic Circuits (AAAI 2024 · similarity 0.81)
- Towards Next-Generation Logic Synthesis: A Scalable Neural Circuit Generation Framework (NeurIPS 2024 · similarity 0.81)
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.
