Research map / Large language models / Agents, code and math
DeepProbLog: Neural Probabilistic Logic Programming
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 symbolic addition rules with a neural digit predicate instead of a CNN that classifies concatenated image pairs | MNIST single-digit addition (T1) · MNIST addition | F1 score | Helped |
| Use symbolic rules to compose digit predictions for multi-digit addition | MNIST multi-digit addition (T2) · Three-digit addition after training on single-digit numbers | accuracy | Helped |
| Use shared convolutional parameters to process each image separately in the CNN baseline | MNIST single-digit addition (T1) · MNIST addition | accuracy | Helped |
| Use DeepProbLog's program sketch rather than Differentiable Forth | Sorting (T4) · Test length 8, training length 4 | accuracy: 49.22 -> 100.0 | Helped |
| Use DeepProbLog's program sketch rather than Differentiable Forth | Sorting (T4) · Test length 64, training length 4 | accuracy: 20.65 -> 100.0 | Helped |
| Use DeepProbLog rather than Differentiable Forth | Sorting (T4) · Training length 4, test length 8 | time to 100% test accuracy: DeepProbLog reaches 100% at training length 4 in 32 s; Differentiable Forth reports -- | Helped |
| Train DeepProbLog on single-digit numbers and apply its reusable digit predicate to multi-digit inputs | MNIST addition (T2) · Three-digit numbers | accuracy | Mixed |
Closest papers by meaning
- DeepStochLog: Neural Stochastic Logic Programming (AAAI 2022 · similarity 0.93)
- Inference and Learning with Model Uncertainty in Probabilistic Logic Programs (AAAI 2022 · similarity 0.89)
- Deep Probabilistic Programming (ICLR 2017 · similarity 0.89)
- Beyond the Grounding Bottleneck: Datalog Techniques for Inference in Probabilistic Logic Programs (AAAI 2020 · similarity 0.89)
- On the Hardness of Probabilistic Neurosymbolic Learning (ICML 2024 · similarity 0.86)
- LogicMP: A Neuro-symbolic Approach for Encoding First-order Logic Constraints (ICLR 2024 · similarity 0.85)
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.