Data

Tools

Indexes

Applications

Experiments

Open Source Science

https://github.com/april-tools/pal

PAL - your probabilistic algebraic layer

https://github.com/april-tools/pal

Science Score: 36.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ○
    CITATION.cff file
  • ✓
    codemeta.json file
    Found codemeta.json file
  • ✓
    .zenodo.json file
    Found .zenodo.json file
  • ○
    DOI references
  • ✓
    Academic publication links
    Links to: arxiv.org
  • ○
    Academic email domains
  • ○
    Institutional organization owner
  • ○
    JOSS paper metadata
  • ○
    Scientific vocabulary similarity
    Low similarity (7.8%) to scientific vocabulary
Last synced: 10 months ago · JSON representation

Repository

PAL - your probabilistic algebraic layer

Basic Info
  • Host: GitHub
  • Owner: april-tools
  • License: mit
  • Language: Jupyter Notebook
  • Default Branch: main
  • Size: 19.4 MB
Statistics
  • Stars: 5
  • Watchers: 2
  • Forks: 0
  • Open Issues: 0
  • Releases: 0
Created over 1 year ago · Last pushed 11 months ago
Metadata Files
Readme License

README.md

PAL - A Probabilistic Neuro-symbolic Layer for Algebraic Constraint Satisfaction

Python application

This repo contains the code for PAL, a probabilistic neuro-symbolic layer for algebraic constraint satisfaction. This is a simplified implementation that focuses on the spline-case.

Check out the paper: https://arxiv.org/abs/2503.19466v1

Example Prediction

This is an example prediction of PAL on the Constrained Stanford Drone Dataset (https://github.com/april-tools/constrained-sdd). We predict a probability distribution over the future trajectory while guaranteeing constraint-satisfaction.

Example image

Citation

Leander Kurscheidt, Paolo Morettin, Roberto Sebastiani, Andrea Passerini, Antonio Vergari, A Probabilistic Neuro-symbolic Layer for Algebraic Constraint Satisfaction, arXiv:2503.19466

Installation

Just clone it and run: bash ./setup.sh And you're ready to go!

Constrained Stanford Drone Dataset

We provide an example script how to train a simple MLP on the constrained SDD-dataset. A model can be trained like this:

bash python pal/training/train_mlp_sdd.py --epochs 10 --init_last_layer_positive --seed 1744909132

This should result in a (mean) test log-likelihood of -1.9149.

GASP!

The dependency was added via subtree from https://github.com/april-tools/gasp.git into pal/wmi/gasp! update via: bash git subtree pull --prefix pal/wmi/gasp https://github.com/april-tools/gasp.git main --squash push via: bash git subtree push --prefix pal/wmi/gasp https://github.com/april-tools/gasp.git main

Owner

  • Name: april-tools
  • Login: april-tools
  • Kind: organization

GitHub Events

Total
  • Watch event: 2
  • Push event: 3
  • Public event: 1
  • Pull request event: 1
Last Year
  • Watch event: 2
  • Push event: 3
  • Public event: 1
  • Pull request event: 1

Dependencies

.github/workflows/python-app.yml actions
  • actions/checkout v4 composite
  • actions/setup-python v3 composite
pal/wmi/gasp/setup.py pypi
  • Explicitly *
  • Install *
  • Main *
  • PySMT >=0.9.6.dev53
  • networkx *
  • numpy *
  • pytest *
  • pytest-runner *
  • sympy >=1.13
  • test *
  • torch >=2.6.0
  • wmipa *
requirements.txt pypi
  • PySMT >=0.9.6.dev53
  • constrained-sdd >=0.1.5
  • frozendict *
  • jupyterlab *
  • matplotlib *
  • networkx *
  • numpy >=2.2.4
  • package_root *
  • pytest >=8.3.5
  • requests >=2.32.3
  • scikit-learn >=1.6.1
  • sympy >=1.13
  • torch >=2.6.0
  • tqdm *